A simple method for the fabrication of WO3 films with electrochromic and photocatalytic properties

Thin Solid Films 573 (2014) 6–13

Contents lists available at ScienceDirect

Thin Solid Films journal homepage: www.elsevier.com/locate/tsf

A simple method for the fabrication of WO3 films with electrochromic and photocatalytic properties G. Leftheriotis a,⁎, M. Liveri a, M. Galanopoulou a, I.D. Manariotis b, P. Yianoulis a a b

Renewable Energy and Environment Lab, Physics Department, University of Patras, Rion 26500, Greece Department of Civil Engineering, University of Patras, Rion 26500, Greece

a r t i c l e

i n f o

Article history: Received 30 January 2014 Received in revised form 27 October 2014 Accepted 27 October 2014 Available online 4 November 2014 Keywords: Tungsten oxide Electrochromism Photocatalysis Spray deposition Precursor aging

a b s t r a c t This work proposes a simple and inexpensive method for the deposition of WO3 films with electrochromic and photocatalytic properties. It consists of preparing a precursor solution of WO3 powder in H2O2, allowing the precursor to age for a period of 1 to 2 months and then spraying it on an appropriate substrate. Changes in the crystallinity of the WO3 particles within the precursor were observed during aging. They were taken advantage of for the preparation of homogeneous WO3 films of optical quality. The films thus prepared are amorphous, stable, with good adherence on the substrate. They exhibit favorable electrochromic properties, such as a high Li ion diffusion coefficient (of about 5 × 10−10 cm2/s) and reversible coloration with efficiencies of 132 cm2/C at 389 nm, 48.9 cm2/C at 534 nm and 52.4 cm2/C at 800 nm. These films were also used as photo-anodes for the photocatalytic degradation of methylene blue (MB) in a NaOH aqueous solution with promising results. Effective degradation of MB was observed with the lowest half-life time of 55 min recorded for a NaOH concentration of 0.8 M. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Tungsten oxide is a compound that has attracted considerable research interest lately, due to its properties that include electrochromism, photochromism, gas sorption and photocatalysis [1,2]. Many applications have been envisaged for it, such as smart windows, switchable mirrors, displays and gas sensors, some of which have already been commercialized [3,4]. WO3 thin films have been developed in the past with the use of physical vapor deposition (thermal, electron beam gun evaporation and sputtering [1,2]) and by chemical methods such as sol–gel, spin coating, electrodeposition and spray pyrolysis [5–8]. Of these, chemical methods are straightforward, versatile and require less expensive equipment. They yield a large variety of film morphologies depending on the pre- and post-deposition conditions, starting materials, solution chemistry [8–11] and thermal treatment of the films [8,12]. They are also well suited for coating of large areas [4]. For the preparation of tungsten oxide films with chemical methods, various starting materials can be used, such as tungsten hexachloride (WCl6) or tungsten oxytetrachloride (WOCl4) for the formation on chloro-alcoxides [7], metallic tungsten (W), tungsten carbide (WC) or sodium tungstate (Na2WO4) that can be diluted in aqueous solutions of hydrogen peroxide to form peroxotungstic acid [7,13–15]. Of these, ⁎ Corresponding author at: Physics Department, University of Patras, 26500 Patras, Greece. E-mail address: [email protected] (G. Leftheriotis).

http://dx.doi.org/10.1016/j.tsf.2014.10.087 0040-6090/© 2014 Elsevier B.V. All rights reserved.

chlorides are reactive and corrosive and need special care in their treatment, while the formation of peroxotungstic acid is a strong exothermal reaction and requires continuous cooling. A much simpler route is the dilution of tungsten oxide in H2O2 that has been used in the past by our group for the preparation of opaque, nanostructured WO3 films [16]. In the present paper, the WO3–H2O2 method is further advanced by allowing the aforementioned precursor solution to age prior to deposition. The aged precursor solution is then sprayed on appropriate substrates. Aging of a different precursor was found to significantly affect the resulting film properties in electrodeposited WO3 films, as was reported in our earlier work [17,18]. Here the aging effect is taken advantage of in order to prepare homogeneous films of optical quality. It was found that aging of the precursor for up to 2 months leads to the deposition of homogeneous WO3 films with favorable electrochromic and photocatalytic properties. 2. Experimental details 2.1. Precursor solution For the preparation of the precursor solution, 0.36 g of WO3 powder (Aldrich, 99.995%) was dissolved into 20 mL of 15% H2O2. With vigorous agitation a yellow-greenish solution was obtained. The solution was unstable and precipitation occurred when the agitation stopped. Heating at 80 °C for about half an hour was used to increase the viscosity of the solution. The solution was aged by storage at room temperature in a sealed bottle.

G. Leftheriotis et al. / Thin Solid Films 573 (2014) 6–13

2.2. WO3 film deposition The precursor solution was used for the growth of WO3 films either immediately after its preparation or after aging which varied from 1 to 4 months. Low-cost equipment were used for the spray deposition: a hand-held electrical air compressor to create suction, a spray nozzle (airbrush), a hot plate for placing the substrates and a magnetic stirrer for mixing the precursor solution prior to spraying. A schematic representation of the spray deposition set-up appears in Fig. 1a. A commercial SnO2:F coated glass product with the trade name Kglass was used as substrate. It is produced by spray pyrolysis with a thickness of 200 to 500 nm [19]. K-glass pieces 2.5 cm × 4 cm in size were used. In spray deposition, film thickness and homogeneity can be controlled by adjustment of numerous parameters such as nozzle-tosubstrate distance, spreading of the solution jet, substrate temperature and number of repetitions. In the present work the nozzle-to-substrate distance was kept constant to 21 cm, the substrate temperature was kept to 80 °C (to avoid “boiling” of the solution upon contact with the glass) and the thickness was varied by spraying on the same substrate more than once. Each spray pulse had a duration of 1 s and a solid film was allowed to form (by drying of the precursor) before the next spray pulse was effected. The precursor drying took a few minutes for every spray pulse.

7

Scanning electron microscopy (SEM) pictures of the films were taken using a JEOL 6300 microscope at 20 kV. In order to assess the electrochemical properties of the films, cyclic voltammetry and Galvanostatic Intermittent Titration (GITT) were performed using the following instruments: a potentiostat–galvanostat (AMEL, model 2053), a function generator (AMEL, model 586) and a noise reducer (AMEL NR 2000). The electrochemical cell used for these experiments has already been described elsewhere [20]. This cell allows for the insertion of a Pt wire serving as the reference electrode, close to the WO3/K-glass (working electrode). A K-glass piece was used as counter electrode. The cell was sealed by o-rings and filled with the 1 M LiClO4-PC liquid electrolyte. During the cyclic voltammetry tests, the potential between the working (WO3/K-glass) and the reference (Pt) electrode was varied linearly in the range (−0.8 V, +1.4 V) at a rate of 20 mV/s. During the GITT experiments, a series of 100 square current pulses with amplitude of 0.2 mA and duration of 5 s were fed into the electrochemical cell, each followed by an equilibration period of 30 s. The potential between working and reference electrodes (corresponding to the electromotive force of the film) versus the inserted charge density was measured. The chemical diffusion coefficient of Li+ was calculated using the formula [18]: D¼

  4L2 ΔEs 2 πτ ΔEt

ð1Þ

2.3. Instrumentation and experimental techniques The thickness of the films was measured using an Ambios XP-1 profilometer. Step height measurements between film-coated and uncoated areas were conducted to determine film thickness. This instrument was also used to assess the surface morphology and roughness of the films. Transmission electron microscopy (TEM) pictures of the dried precursor were taken using a JEOL JEM 2100 microscope at 160 kV. A drop of the precursor solution was placed on a carbon coated copper substrate (grid) and was allowed to dry prior to the measurements.

where ΔEs is the change in the steady-state voltage as a result of the current pulse, ΔEt is the total change in overvoltage during the pulse, τ is the duration of each pulse and L is the film thickness. Eq. (1) can be applied to our electrochemical systems provided that the following conditions are observed [21]: (a) Negligible change in the WO3 molar volume with Li+ intercalation. (b) Small current values to justify the approximation dE/dx ≈ ΔE/Δx, i.e. the substitution of a first derivative by a finite difference. (c) During the current pulse the overvoltage E is proportional to t1/2. (d) τ ≪ L2/D. The experimental parameters were set so as to meet these conditions. The “intercalation parameter” x in the LixWO3 films was calculated using the following expression [21]:   Iτ Npulse qe x¼  NA ρ PD S L M

Fig. 1. Schematic representation of the experimental set-up. (a) Spray deposition of WO3, (b) photocatalysis experiments.

ð2Þ

where I and τ are the current and the duration of each pulse respectively, qe is the elementary charge, Npulse is the pulse number, NA the Avogadro number and M, ρ, PD, S, and L are the molecular weight, bulk density, packing density, area and thickness of the WO3 film, respectively. Electrochromic (EC) devices of the form K-glass/WO3/liquid electrolyte/K-glass were fabricated, incorporating the prepared WO3 films. The liquid electrolyte used in all experiments was 1 M lithium perchlorate dissolved in propylene carbonate (LiClO4-PC). In order to construct the electrochemical cells, two pieces of K-glass (a plain one and one with the deposited tungsten oxide film) were used, each having dimensions of 2.5 cm by 4 cm. The two pieces of glass were arranged facing each-other and a cavity was formed between them using silicone. The cavity was then filled with the 1 M LiClO4-PC liquid electrolyte and sealed with silicone [17]. Electrical contacts were created at the edge of each piece of glass using copper adhesive tape. The electrochromic devices thus fabricated were also subjected to a galvanostatic coloration process. Certain amounts of electric charge,

8

G. Leftheriotis et al. / Thin Solid Films 573 (2014) 6–13

ranging from 2 to 31 mC/cm2, were gradually fed into the devices. The transmittance spectra of the devices during different coloration stages were then recorded using a PerkinElmer Lambda 650 UV/VIS Spectrometer. The coloration efficiency was calculated from the change in the optical density and the charge density fed into each device during coloration. Specifically, the change in optical density (ΔOD) and coloration efficiency (CE) were calculated from the following relations [1]: ΔOD ðλÞ ¼ log½T bleached ðλÞ=T colored ðλÞ

ð3Þ

CE ðλÞ ¼ ΔOD ðλÞ=Q

ð4Þ

with Tbleached and Tcolored the cell transmittance in the bleached and colored state respectively and Q the intercalated charge density, which corresponds to the ratio of the inserted charge over the device active area. The Haze number (H) of the devices was estimated by measurement of the total and diffuse transmittance spectra {TT(λ) and DT(λ) respectively} using a PerkinElmer Lambda 35 UV/VIS spectrometer equipped with an integrating sphere. 2.4. Photocatalysis experiment The photocatalysis arrangement was placed in a dark room and a schematic representation is illustrated in Fig. 1b. An appropriate volume of methylene blue (MB) solution (20 mg/L) was transferred into a 250mL Pyrex glass beaker. The anode and cathode electrodes were connected with a DC power supply, and were immersed into the solution. The DC voltage was adjusted to the maximum possible value that did not lead to electrode activity in the dark. It was found to be about 3 V. The applied electrical potential is necessary in order to drive the photoelectrons from the anode to the cathode and thus avoid carrier recombination. This lay-out of the photo-electrochemical cell is also known as “cell with external bias” [30]. A high pressure mercury lamp (Osram, 125 W) was connected to an appropriate transformer. The lamp was placed horizontally and aligned at the same level with the liquid in the beaker in order to irradiate the WO3 film. The incident light intensity has been measured in the range of 0.2 μm to 20 μm with the use of a Melles Griot 13PEM001 power meter. It was found to be about 250 W/m2. A platinum film on a SnO2:F coated glass was used as cathode. It was prepared by electrodeposition of a hexachloroplatinic acid solution (H2PtCl6), as described in [22]. The cathode electrode was placed at right angles to the photo-anode to avoid direct exposure to the UV source. A working solution of 20 mg/L methylene blue was used in all experiments, dissolved in NaOH solutions of 0.01, 0.1, 0.4, and 0.8 M in order to examine the effect of hydroxyl concentration. During each experiment, 4 mL samples were taken every 30 min. Samples were collected with a glass pipette and stored in glass vials. The concentration of methylene blue was deduced by absorbance measurements at 590 nm, by a cuvette mode UV–VIS spectrophotometer (U1100, Hitachi). The temperature of the liquid was measured with a digital thermometer equipped with a thermistor (Oakton Temp 5 Acorn Series, Eutech Instruments Ltd., Singapore). The experiments were conducted at room temperature. 3. Results and discussion 3.1. Precursor solution Transmission electron microscopy pictures of the precursor solution appear in Fig. 2. The as prepared solution of Fig. 2a comprises of grains a few hundred nm in size. The observed grain size is much smaller than that of the WO3 powder used (20 μm average particle size). This indicates that the powder decomposes to smaller particles upon dilution in the peroxide solution. On the other hand, in the 2.5 months aged solution, several particles are joined together, as can be seen in Fig. 2c. The

a

100 nm

b

10 nm-1

c

100 nm

d

10 nm-1

Fig. 2. TEM images and SAE diffractograms of the precursor solution. (a), (b) as prepared, (c), (d) 2.5 months aged.

Selected Area Electron (SAE) diffractograms of Fig. 2b and d give more information on the crystallinity of the WO3 particles: In the asprepared state the grains appear crystalline with distinct diffraction spots [23]. On the other hand, the aged solution grains appear amorphous as revealed by the characteristic diffuse rings that dominate the SAED image [23]. Some degree of crystallinity is present in the aged grains, as some direction spots are still present. Thus we can conclude that the crystalline grains of the as prepared precursor solution are slowly transformed to amorphous by aging. This effect has been verified by the WO3 films that were prepared, as will be discussed next. 3.2. Properties of the spray-deposited WO3 films 3.2.1. Morphology of the films In Fig. 3a a SEM image is shown of a WO3 film deposited by the spray method on SnO2:F coated glass. The film was prepared immediately after the synthesis of the precursor solution. It consists of spheroid grains, 300 nm to 700 nm in diameter, similar to those shown in Fig. 2a. It can therefore be concluded that the WO3 powder grains dissolved in the H2O2 solution, are “transferred” to the film without any modification in their morphology. The films thus prepared have a yellowish color. They do not adhere well on the substrate, as they can be removed from it by adhesive tape (e.g. they do not withstand the “Scotch tape test”). Compared with films developed from an identical precursor by drop casting [16], they have a similar morphology, as shown in Fig. 1 of [16]. However the films developed in [16] are more robust, as they can withstand the Scotch tape test. This is probably due to the fact that drying of the drop-casted films takes longer (1 h at 80 °C), thus facilitating the formation of bonds among the grains and the substrate. The crystallinity of such films can be assessed by electrochemical testing, and in particular by the GITT method. It is known [24] that the presence of plateaus in the electromotive force (EMF) curves deduced from GITT experiments reveals crystallinity of the tested samples. On the other hand, a monotonic EMF is typical of amorphous materials [17]. The GITT results appear in Fig. 4. The WO3 film tested appears crystalline as its EMF curve is not monotonic. The Li ion diffusion coefficient calculated by Eq. (1) exhibits two local maxima at about 0.0025 and 0.0125 of the intercalation parameter x (Eq. (2)). The crystalline nature of the films is due to the crystallinity of the grains in the precursor solution, as shown in Fig. 1b.

G. Leftheriotis et al. / Thin Solid Films 573 (2014) 6–13

a

9

b

1μm

2 μm

10 μm

10 μm

v

Fig. 3. SEM images of WO3 films on SnO2:F coated glass. (a) Immediately after the formation of the precursor solution. (b) After annealing at 400 °C for 1 h.

0.1 0.0 -0.1 -0.2 -0.3 -0.4 -0.5 -0.6 -0.7 -0.8 -0.9 -1.0 -1.1 -1.2 -1.3

1.E-08

1.E-09

1.E-10

EMF as prepared 1.E-11

EMF 400 deg D as prepared

Diffusion Coefficient (cm 2/s)

EMF (Volt)

3.2.2. Effect of annealing To assess the effect of annealing on the WO3 films, samples deposited immediately after the synthesis of the precursor solution were subjected to heating at 400 °C for 1 h in the atmosphere. The results appear in Figs. 3b and 4. The SEM image of Fig. 3b reveals the formation of rods of various sizes after annealing, thus, it can be concluded that heating has caused re-crystallization of the films. Apart from the observed morphological modification, the film structure is also expected to change. It is well known that annealed tungsten oxide films undergo a series of phase transitions in sequence from monoclinic to orthorhombic at 350 °C and then to hexagonal at 500 °C [25]. An indication of the expected alteration in the film crystal structure appears in Fig. 4, where a change of the EMF curve is evident. This in turn affects the diffusion coefficient that exhibits one maximum at x = 0.0065 (different than the two maxima of the as prepared film) and then decreases to values as low as 2 × 10−11 cm2/s. Such low values of D are indicative of films with poor electrochromic behavior. The electrochromic performance of the annealed film was assessed by cyclic voltammetry experiments. As can be seen in Table 1, the intercalated charge density (Qin), as calculated by numerical integration of the voltammograms [18], is considerably lower for the annealed film, compared to that of the same film prior to annealing. The same applies for the cathodic peak current density (Jmax). It can therefore be concluded that although re-crystallization of the WO3 films during annealing gives an interesting rod-like morphology, it is detrimental to their electrochromic performance. For that reason, annealing was not pursued any further.

D 400 deg 0

0.0025

0.005

0.0075

0.01

0.0125

1.E-12 0.015

x in LixWO 3 Fig. 4. EMF and Li+ chemical diffusion coefficient D versus the intercalation parameter x (in LixWO3) for a WO3 film deposited immediately after the precursor formation and for the same film annealed at 400 °C for 1 h.

3.2.3. Effect of the precursor aging It has already been shown that storage causes alterations to the properties of the precursor solution. WO3 films were deposited using aged solutions to investigate the aging effects. The results are presented next. The SEM images of Fig. 5 show WO3 films prepared by aged precursors. In contrast with the granular nature of films of Fig. 3a, the films from the 1 and 2 months aged precursors appear continuous with a small degree of grains incorporated in the structure. Another feature characteristic of these films is a network of cracks, caused probably during drying of the precursor. In the 2 months aged precursor, the cracks are narrower. The picture changes once more for the 2.5 months aged solution. The film of Fig. 5c is not continuous and grains of irregular shape are formed. It is therefore clear that aging of the precursor for up to 2 months leads to an improvement of the WO3 film morphology. This can also be verified by the surface profiles of these films as shown in Fig. 6. Indeed, as can be seen therein, only the films coming from 1 month and 2 months aged precursors have a continuous surface profile, with reduced roughness, as opposed to the films with the as prepared and 2.5 months aged precursor, which appear highly irregular and grainy. The films coming from aged precursors up to 2.5 months are pale white in color and adhere well to the substrates. They withstand the Scotch tape test and can only be scratched away by a sharp metal object. Further aging of the precursor was also tried. At 4 months of aging, the films prepared resembled more to the starting material, having the form of a yellowish powder with extremely poor adherence to the substrate. Moderate blowing can remove the films from the SnO2:F coated glass. Another phenomenon was also observed during the spray process: The temperature of the aged precursor was found to increase when the solution was stirred prior to spraying. The solution temperature was thus raised up to 60 °C, revealing that stirring gave rise to an exothermal reaction. A possible explanation of this effect could be the for2− , in mation of the dimer W2O2− 11 or [(O)W(O2)2(O)(O2)2W(O)] which tungsten is in a (+VI) oxidation state and (O2) denotes a peroxide ligand [14,15]. This anion is normally formed by the dissolution of tungsten metal in concentrated hydrogen peroxide and the reaction is strongly exothermal [14]. In that case, the resulting WO3 films prepared by electrodeposition are amorphous, similar to the “aged precursor” films presented herein [14,17]. The morphological transformation of the aged precursor has been verified by TEM (Fig. 1d) and by the fact that the resulting WO3 films are distinctively different. It is possible that this particular effect takes place during aging of the precursor and it is speeded-up by stirring. Further experimental work is necessary to fully understand the reactions that take place in the solution during aging and to find ways to speed-up the procedure. The GITT results for films grown from aged precursors appear in Fig. 7. As can be seen therein, the EMF curves appear monotonic and

10

G. Leftheriotis et al. / Thin Solid Films 573 (2014) 6–13

Table 1 Properties of the WO3 films. Precursor/treatment

Average thickness (μm)

Jmax (mA cm−2)

Jmax/thickness (mA cm−2 μm−1)

Intercalated charge density Qin (mC cm−2)

As prepared As prepared/400 °C 1 month aged 2 months aged 2.5 months aged

1.69 1.69 1.86 1.97 1.12

−1.36 −0.18 −0.81 −1.33 −1.16

−0.80 −0.11 −0.44 −0.68 −1.04

−17.45 −3.55 −18.79 −36.04 −40.53

we can therefore conclude that the films are predominantly amorphous. This finding agrees with the SAE Diffractogram of the 2.5 month precursor, shown in Fig. 2d. The corresponding Li+ diffusion coefficients shown in Fig. 7 do not exhibit dramatic changes with increasing

a

5 μm

60 μm

b

intercalation parameter x. Of the WO3 films prepared with aged precursors, the 2 months one exhibits the highest diffusion coefficient, of about 5 × 10−10 cm2/s, a value suitable for electrochromic applications. The electrochemical response of the WO3 films prepared by the spray method was also assessed by cyclic voltammetry experiments. The results appear in Fig. 8. It can be seen therein that the curves pertaining to the films grown from aged precursors are different than those of the film from the as prepared solution. The latter exhibits two positive peaks (at about + 350 mV and + 790 mV), typical of WO3 films fabricated by wet methods [16–18]. This feature is lacking in the films grown from aged precursors that exhibit features similar to evaporated WO3 films [26]. Furthermore, differences can be observed among the films from aged precursors. Increasing aging time causes all the desirable features of the resulting films to improve. Indeed, as shown in Table 1, the peak cathodic current per unit thickness (Jmax/thickness) and the intercalated charge density (Qin) increase with aging time increasing. These changes are due to the transition of the films from multi-crystalline to amorphous, as it is well known that the amorphous phase is more open and favors intercalation of Li+ cations [1]. From the EC applications point of view, optical quality and homogeneity of the films are also crucial, thus, the films resulting from the 2 months aged precursor appear the most promising for EC devices: They combine improved morphology with favorable electrochemical properties. In the following section the testing of such an EC device is presented. 3.3. Applications of the WO3 films

5 μm

60 μm

c

1μm

60 μm

Fig. 5. SEM images of WO3 films on SnO2:F coated glass using precursors aged for (a) 1 month, (b) 2 months, (c) 2.5 months.

3.3.1. Electrochromic devices A WO3 film fabricated from the 2 months aged precursor was incorporated in an electrochromic device of the form K-glass/WO3/liquid electrolyte/K-glass, as explained in Section 2.3. A picture of the device in the bleached and colored state appears in Fig. 9. In Fig. 10a the transmittance spectra of the device are shown, at various coloration stages. The coloration-bleaching process was found to be reversible, and the device was cycled periodically (once every day) for 70 days with no indication of degradation. The coloration efficiency CE was calculated with the use of Eq. (4) and its maximum values were found to be 132 cm2/C at 389 nm, 48.9 cm2/C at 534 nm and 52.4 cm2/C at 800 nm. These values of CE compare well with those of WO3 films prepared by other techniques: evaporation [26], electrodeposition [17,18], and spray pyrolysis [27]. CE is particularly high at 389 nm, due to the fast coloration observed at lower wavelengths, especially in the initial stages of the process. This is another interesting property of the fabricated films, not found in typical electrochromic devices. The reason for this feature is not clear and should be the subject of further investigation. The only disadvantage of the fabricated EC device is its rather low transmittance in the bleached state, as can be seen in Fig. 10a. In order to assess whether the low transmittance of the devices is due to light scattering, we have estimated their transmission Haze number (H) in the bleached state, defined as the ratio of Diffuse to Total Transmittance (DT/TT%) following the ASTM test method D1003, described in [31]. The results appear in Fig. 10b in the form of wavelength spectra. If follows that H is large throughout the measured spectrum (above 40%). Further

G. Leftheriotis et al. / Thin Solid Films 573 (2014) 6–13

11

9000 As prepared

8000

Height (nm)

7000 6000 Aged for 1 month

5000

Aged for 2.5 months

Aged for 2 months

4000 3000 2000 1000 0 0

500

1000

1500

2000

2500

3000

3500

Horizontal distance (µm) Fig. 6. Surface profiles of WO3 films deposited with as prepared and aged precursors.

improvements in the spray deposition method could yield better control of the resulting film thickness and morphology, leading to more transparent films. Nevertheless, the films developed herein could be suitable as low cost alternatives, in translucent electrochromic windows and in electrochromic display applications.

NaOH concentrations. According to this theory, the coverage ratio θ is defined as [28]:

3.3.2. Photocatalysis of methylene blue A 4 μm thick WO3 film, prepared by the 2 months aged precursor, was used as the anode in the photocatalysis apparatus described in Section 2.3. The results appear in Fig. 11, as the normalized concentration (C/C0) of MB versus time, for electrolytes with different concentrations of NaOH, ranging from 0.01 M to 0.8 M. The curves that fit best each set of measurements also appear in Fig. 11. It can be observed that with increasing NaOH concentration the photocatalysis procedure speeds-up. It can also be observed, that low NaOH concentrations (namely 0.01 M and 0.1 M) obey zero order kinetics. For the highest NaOH concentration (0.8 M) the degradation of MB follows first-order kinetics, while for 0.4 M NaOH, MB degradation can be divided into two parts, one following zero-order, and the other one following first order. The inset of Fig. 11 depicts the same curves normalized by the half-life time t1/2 (e.g. C/C0 vs t/t1/2) for all NaOH concentrations apart from 0.01 M, for which the relative MB concentration was not halved during the measured interval. It is clear that the normalized plots swarm around the same curve, therefore it can be concluded that the same trend is followed under different conditions. These observations can be explained by the Langmuir–Hinshelwood theory with the assumption of a limitation in adsorption sites at lower

with nads the number of adsorbed molecules, n0 the number of the available adsorption sites, C the MB concentration and K the constant of MB adsorption on the WO3 surface at equilibrium. Thus, for high values of KC (KC ≫ 1), θ is close to unity, as all the available sites contain adsorbed MB molecules. The rate of the MB degradation r is given by [28,29]:

EMF 1m aged

EMF 2m aged

EMF 2.5m aged

D 1m aged

D 2m aged

D 2.5m aged

nads KC ¼ n0 1 þ KC

r¼−

ð5Þ

dC KC ¼k dt 1 þ KC

ð6Þ

where k is the true rate constant. For low values of KC, (KC ≪ 1) the coverage ratio of Eq. (5) is low and Eq. (6) gives an exponential time dependence of the MB concentration. With increasing KC values, saturation of the available adsorption sites occurs, and the evolution of the reaction slows down, as can be inferred by the half-life time t1/2, derived by integration of Eq. (6): t 1=2 ¼

C0 ln 2 þ 2k kK

ð7Þ

2 where C0 is the initial concentration, and ln kK is the half life time of the exponential relation (when KC ≪ 1 holds). It follows from Eq. (7) that kt1/2 is constant for constant C0 and K, thus the non-dimensional form of Eq. (6) normalized by C0 and the half life time t1/2, yields the same

1.0 1.E-08

0.8

1.E-09

0.6 0.4

1.E-10

0.2 1.E-11

0.0 -0.2

1.E-12

-0.4 -0.6

0.5

J (mA/cm2)

1.0

Diffusion Coefficient (cm2/s)

1.2

EMF (Volt)

θ¼

0.0

-0.5 As prepared 1 month aged -1.0

2 months aged 2.5 months aged

-0.8

1.E-13 0

0.02

0.04

0.06

0.08

0.1

x in LixWO3 Fig. 7. EMF and Li+ chemical diffusion coefficient D versus the intercalation parameter x (in LixWO3) for WO3 films deposited from aged precursor solutions.

-1.5 -1000

-500

0

500

1000

1500

V (mV) Fig. 8. Cyclic voltammograms of WO3 films deposited with as prepared and aged precursors.

12

G. Leftheriotis et al. / Thin Solid Films 573 (2014) 6–13

Fig. 9. Pictures of the fabricated electrochromic device in the bleached and colored state.

result for different NaOH concentrations, as was observed in the inset of Fig. 11. In the present photocatalysis study C0 is constant, equal to 20 mg/L. Thus, the initial KC0 value is dictated by MB adsorption on the oxide surface, which in turn depends on the WO3 surface charge. Indeed, it is well known that the surface charge of an oxide (i.e. titania) is affected by the presence of hydroxyl groups [29]. In an alkaline environment, with pH values higher than the point of zero charge, the oxide surface is negatively charged. In the present system, the higher the NaOH concentration, the higher the WO3 surface charge. It is known that methylene blue has a cationic configuration, thus its adsorption on the photoanode is favored in alkaline solutions [29]. Furthermore, the initial step of MB degradation has been ascribed to the cleavage of the bonds of the C\S+_C functional group in the presence of the OH• radicals that attack that group, when MB is in direct coulombic interaction with photo-anode surface [29]. Thus, at low NaOH concentrations few adsorption sites exist on the WO3 surface, KC0 is high and the MB degradation progresses slowly. The minimum value of MB half-life is about 55 min, for the highest NaOH concentration (0.8 M), as can be seen in Fig. 11. The photo-catalytic performance of the fabricated WO3 films was compared with that of equivalent mesoporous TiO2 films, 4 μm thick, with grain size of about 40 nm, prepared by the “doctor blade” method, as described in [32]. The same lay-out has been used for the degradation of MB, to obtain a fair comparison. TiO2 was found to exhibit a similar behavior as that of WO3 with higher half-life times: 193 min for 0.1 M of NaOH (about 10% higher than that of WO3) and 68 min for 0.8 M of NaOH (about 20% higher than that of WO3). Therefore, the fabricated WO3 layers exhibit higher photocatalytic activity than the TiO2 films. This was expected as WO3 has a lower energy gap than TiO2 [30].

2 mC/cm2

Bleached 15mC/cm2 mC/cm2 15

5 mC/cm2

These preliminary results show that the WO3 films prepared within possess promising photocatalytic properties, and are appropriate for pollutant degradation applications. 4. Conclusions A simple and inexpensive method for the deposition of WO3 films has been devised. It consists of preparing an aqueous precursor solution of WO3 powder in H2O2, allowing the precursor to age for a period of 1 to 2 months and then spraying it on an appropriate substrate. Changes in the crystallinity of the WO3 particles within the precursor were observed during aging. The particles appear to change from predominantly multi-crystalline to predominantly amorphous. Annealing of the WO3

b

31mC/cm2 mC/cm2 31

TT

DT

Haze number 100

60

20

18

50

80

16 40

60

12

T (%)

T (%)

14

10

30 40

8

20

6 4

Haze number (%)

a

Fig. 11. Photocatalysis of methylene blue: normalized concentration of MB vs time, for electrolytes with different NaOH concentrations. Inset: Normalized plots of relative concentration vs t/t1/2.

20

10

2 0 400

500

600

700

Wavelength (nm)

800

0 400

500

600

700

800

900

0 1000

Wavelength (nm)

Fig. 10. (a) Transmittance spectra of a fabricated electrochromic device at various coloration stages. (b)Total (TT) and diffuse (DT) transmittance spectra of an electrochromic device in the bleached state.

G. Leftheriotis et al. / Thin Solid Films 573 (2014) 6–13

films at 400 °C for 1 h in the atmosphere causes morphological and structural alterations to the films that render them unsuitable for electrochromic applications. The WO3 films prepared with the use of a 2 months aged precursor are amorphous, stable, with good adherence to the substrate. They exhibit favorable electrochromic properties, such as a high Li ion diffusion coefficient (of about 5 × 10− 10 cm2/s) and reversible coloration with efficiencies of 132 cm2/C at 389 nm, 48.9 cm2/C at 534 nm and 52.4 cm2/C at 800 nm. These films were also used as photo-anodes for the photocatalytic degradation of methylene blue (MB) with promising results: Effective degradation of MB was observed with the lowest half-life time of 55 min recorded for a concentration of 0.8 M NaOH. Acknowledgments G. Leftheriotis and P. Yianoulis acknowledge the financial support from the “Karatheodori 2010–2013” research grants of Patras University (D.212). References [1] C.G. Granqvist, Handbook of Inorganic Electrochromic Materials, first ed. Elsevier Science, Amsterdam, 1995. [2] K. Deb Satyen, Opportunities and challenges in science and technology of WO3 for electrochromic and related applications, Sol. Energy Mater. Sol. Cells 92 (2008) 245. [3] R. Baetens, B.P. Jelle, A. Gustavsen, Properties, requirements and possibilities of smart windows for dynamic daylight and solar energy control in buildings: a state-of-the-art review, Sol. Energy Mater. Sol. Cells 94 (2010) 87. [4] A. Kraft, M. Rottmann, Properties, performance and current status of the laminated electrochromic glass of Gesimat, Sol. Energy Mater. Sol. Cells 93 (2009) 2088. [5] S.R. Bathe, P.S. Patil, Electrochromic characteristics of fibrous reticulated WO3 thin films prepared by pulsed spray pyrolysis technique, Sol. Energy Mater. Sol. Cells 91 (2007) 1097. [6] J.M. Ortega, A.I. Martınez, D.R. Acosta, C.R. Magana, Structural and electrochemical studies of WO3 films deposited by pulsed spray pyrolysis, Sol. Energy Mater. Sol. Cells 90 (2006) 2471. [7] J. Livage, D. Ganguli, Sol–gel electrochromic coatings and devices: a review, Sol. Energy Mater. Sol. Cells 68 (2001) 365. [8] M. Deepa, A.K. Srivastava, T.K. Saxena, S.A. Agnihotry, Annealing induced microstructural evolution of electrodeposited electrochromic tungsten oxide films, Appl. Surf. Sci. 252 (2005) 1568. [9] B. Yang, H. Li, M. Blackford, V. Luca, Novel low density mesoporous WO3 films prepared by electrodeposition, Curr. Appl. Phys. 6 (2006) 436. [10] Q. Zhong, K. Colbow, Electrochromic properties of cesium tungstate with pyrochlore structure, Thin Solid Films 205 (1991) 85. [11] Z. Yu, X. Jia, J. Du, J. Zhang, Electrochromic WO3 films prepared by a new electrodeposition method, Sol. Energy Mater. Sol. Cells 64 (2000) 55.

13

[12] A.K. Srivastava, M. Deepa, S. Singh, R. Kishore, S.A. Agnihotry, Microstructural and electrochromic characteristics of electrodeposited and annealed WO3 films, Solid State Ionics 176 (2005) 1161. [13] K. Yamanaka, H. Oakamoto, H. Kidou, T. Kudo, Peroxotungstic acid coated films for electrochromic display devices, Jpn. J. Appl. Phys. 5 (1986) 1420. [14] E.A. Meulenkamp, Mechanism of WO3 electrodeposition from peroxy-tungstate solution, J. Electrochem. Soc. 144 (1997) 1644. [15] T. Pauporte, A simplified method for WO3 electrodeposition, J. Electrochem. Soc. 149 (2002) C539. [16] G. Leftheriotis, S. Papaefthimiou, P. Yianoulis, A. Siokou, D. Kefalas, Structural and electrochemical properties of opaque sol–gel deposited WO3 layers, Appl. Surf. Sci. 218 (2003) 275. [17] G. Leftheriotis, P. Yianoulis, Development of electrodeposited WO3 films with modified surface morphology and improved electrochromic properties, Solid State Ionics 179 (2008) 2192. [18] M. Giannouli, G. Leftheriotis, The effect of precursor aging on the morphology and electrochromic performance of electrodeposited tungsten oxide films, Sol. Energy Mater. Sol. Cells 95 (2011) 1932. [19] T.M. Letcher, J.L. Scott (Eds.), Materials for a Sustainable Future, RSC Publishing, Cambridge, 2012, p. 648. [20] S. Papaefthimiou, G. Leftheriotis, P. Yianoulis, Advanced electrochromic devices based on WO3 thin films, Electrochim. Acta 49 (2001) 2145. [21] G. Leftheriotis, S. Papaefthimiou, P. Yianoulis, Dependence of the estimated diffusion coefficient of LixWO3 films on the scan rate of cyclic voltammetry experiments, Solid State Ionics 178 (2007) 259. [22] G. Syrrokostas, A. Siokou, G. Leftheriotis, P. Yianoulis, Degradation mechanisms of Pt counter electrodes for dye sensitized solar cells, Sol. Energy Mater. Sol. Cells 103 (2012) 119. [23] R. Schäfer, P.C. Schmidt (Eds.), Methods in Physical Chemistry, Wiley-VCH, New York, 2012, p. 806. [24] M. Stromme Mattsson, Cation intercalation in sputter-deposited W oxide films, Phys. Rev. B 58 (1998) 11015. [25] C.V. Ramana, S. Utsunomiya, R.C. Ewing, C.M. Julien, U. Becker, Structural stability and phase transitions in WO3 thin films, J. Phys. Chem. B 110 (2006) 10430. [26] G. Leftheriotis, E. Koubli, P. Yianoulis, Combined electrochromic-transparent conducting coatings consisting of noble metal, dielectric and WO3 multilayers, Sol. Energy Mater. Sol. Cells 116 (2013) 110. [27] P.M. Kadam, N.L. Tarwal, P.S. Shinde, R.S. Patil, H.P. Deshmukh, P.S. Patil, From beads-to-wires-to-fibers of tungsten oxide: electrochromic response, Appl Phys A 97 (2009) 323. [28] J.-M. Herrmann, Heterogeneous photocatalysis: state of the art and present applications, Top. Catal. 34 (2005) 49. [29] A. Houas, H. Lachheb, M. Ksibi, E. Elaloui, C. Guillard, J.-M. Herrmann, Photocatalytic degradation pathway of methylene blue in water, Appl. Catal. B Environ. 31 (2001) 145. [30] P. Lianos, Production of electricity and hydrogen by photocatalytic degradation of organic wastes in a photoelectrochemical cell. The concept of the photofuelcell: a review of a re-emerging research field, J. Hazard. Mater. 185 (2011) 575. [31] H.L. Yu, C.C. Hsiao, W.C. Liu, New apparatus for haze measurement for transparent media, Meas. Sci. Technol. 17 (2006) N29. [32] G. Leftheriotis, G. Syrrokostas, P. Yianoulis, “Partly covered” photoelectrochromic devices with enhanced coloration speed and efficiency, Sol. Energy Mater. Sol. Cells 96 (2012) 86.