Characterization and mechanism analysis of AgBr mixed cuboid WO₃ rods with enhanced photocatalytic activity

Abstract

Cuboid rods of monoclinic WO₃ (m-WO₃) were successfully prepared by employing preoxo-polytungstic acid as the precursor via a hydrothermal method. AgBr/m-WO₃ photocatalysts with varying loadings of AgBr were synthesized and characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM). Photocatalytic degradation of rhodamine B (rhB), methyl orange (MO) and methyl blue (MB) were carried out to evaluate the photocatalytic activity of AgBr/m-WO₃ under visible-light irradiation (λ ≥ 420 nm). Among such catalysts, the AgBr (30 wt%)/m-WO₃ exhibits the highest visible-light-responsible photoactivity. However, these AgBr/m-WO₃ materials gradually lost their photoactivity in the cycling photocatalytic tests. Possible mechanisms for both the enhance photocatalytic activity and deactivation of the AgBr/m-WO₃ were proposed on basis of theoretical speculation and experimental observations.

---

Introduction

Metal–oxide semiconductors may be used in a wide variety of applications including photocatalytic degradation of organic compounds and gas monitoring and alarm systems.^1,2 Materials based on tungsten trioxide (WO₃) have been studied extensively because of their special electronic and optoelectronic properties, which have enormous potential applications in fields ranging from condensed-matter physics to solid-state chemistry,³ such as photo-electrochemical energy conversion,⁴ gas-sensors,⁵ photo-catalysts,⁶ lithium-ion batteries,⁷ solar cells,⁸ electron emitters,⁹ optical storage media.¹⁰ The band gap of WO₃ is situated within the solar spectrum range, making it an important material for photocatalytic activity and photoconductivity.¹¹ However, like other simple binary metal oxides, WO₃ has a deep valence band which is mainly composed of O_2p orbitals. The deep valence band combined with the small band gap results in a low conduction band level, which limits the photocatalyst to react with electron acceptors^12–14 and then increases the recombination of photogenerated electron–hole pairs. This was one of the principles to improve the photocatalytic performance of WO₃ for increasing the efficiency of electron–hole separation. Loading silver halides on the photocatalyst has been proved as effectual approach recently.^16–20 Silver halides (AgX, X = Cl, Br, I) are important photosensitive semiconductors extensively employed in photography field. Under light irradiation AgX can absorb photons to generate electrons and holes. Therefore, it will be a promising way to utilize AgX as photosensitive components in photocatalytic field.^21,22

For the above mentioned applications, the microstructures of photocatalyst also influence the photocatalytic activity significantly. WO₃ has been produced by various synthetic routs such as electrospinning,²³ chemical and physical vapor deposition,²⁴ co-precipitation,²⁵ sol–gel processing,²⁶ thermal processing,²⁷ microwave synthesis²⁸ and solvothermal/hydrothermal methods.²⁹ Herein, the cuboid rods of monocline WO₃ (m-WO₃) has been successful control using peroxo-polytungstic acid as the precursor under hydrothermal conditions. Furthermore, considering that AgX (X = Cl, Br) have highly photosensitive materials, we set out to investigate the photocatalytic performance of AgBr/m-WO₃ and its working mechanism as visible-light-responsive photocatalysts.

Here, we reported the photocatalytic of AgBr/m-WO₃ composite catalyst in the degradation reaction of rhodamine B (rhB), methyl orange (MO) and methyl blue (MB) under the simulated sunlight irradiation. The conduction band (CB) bottom and valence band (VB) tope of m-WO₃ lie below the CB bottom and VB to of AgBr, respectively, which will result in the highly efficient separation of photoinduced electrons and holes. In addition, AgBr/m-WO₃ can easily transform to be Ag/AgBr/m-WO₃ system in the early stage of the photocatalytic reaction, which in the role of Ag nanoparticles. The stability of the photocatalyst was also investigated via the repetition tests. However, AgBr/m-WO₃ gradually lost their photoactivity in the cycling photocatalytic tests. The possible mechanisms for the formation of photocatalysts, enhanced photocatalytic properties and deactivation of the AgBr/m-WO₃ were investigated, respectively.

Experimental

Preparation and characterization

Tungsten powder, hydrogen peroxide, AgNO₃, glycerol and cetyltrimethyl ammonium bromide (CTAB) are of analytical purity from Wako Pure Chemical Industries and used for the experiment without further purification. Deionized water was used throughout this study.

The cuboid rods of m-WO₃ were synthesized using peroxo-polytungstic acid as precursor in hydrothermal reaction according to the previous report with minor changes.³⁰ Typically, finely powdered tungsten (5 g, 12 μm) was slowly added to hydrogen peroxide (20 ml, 30 wt%) in a cold water bath (10 °C) to stabilize the reaction temperature below 30 °C. This produced a clear and colourless peroxo-polytungstic acid solution.

The as-prepared solution was then diluted in deionized water to obtain a tungsten concentration of 0.127 mol L⁻¹. After that, the diluted solution was transferred into a Telfon-lined stainless steel autoclave (60 ml, filling ration 75%), and the autoclave was then sealed and maintained at 180 °C for 4 days. After cooled to room temperature, the as-prepared precipitates were isolated by centrifugation and repeatedly washed with absolute ethanol and deionized water for several times. Finally, the as-obtained precipitates were dried at 40 °C in a drying cabinet for 6 h. Then the cuboid rods of monoclinic WO₃ (m-WO₃) was obtained.

In a typical synthesis, the as-synthesized m-WO₃ (0.5 g) was added to distilled water (30 ml), and the suspension was sonicated for 30 min. The glycerol (5 ml) was added to suspension, and the mixture was sonicated for 30 min. The solution was mixed various amounts of AgNO₃ in order to obtain AgBr/m-WO₃ composite with AgBr weight percentage of 5.0, 10.0, 20.0, 30.0, 40.0. The mixed solution was vigorously stirred for 30 min in the dark prior to adding to 25 ml deionized water containing stoichiometric amounts of CTAB. The resulting suspension was stirred at room temperature for 12 h. All the above processes were performed in the dark. The products were filtered, washed with absolute ethanol and deionized water. Finally, the precipitates were dried at 40 °C for 12 h in vacuum. Then, the AgBr/m-WO₃ composite catalysts were obtained. Fig. 1 describes the detailed preparation process.

Fig. 1 The schematic diagram of the preparation procedure of AgBr/m-WO₃ photocatalyst.

The phase compositions of the AgBr/m-WO₃ samples were determined by X-ray diffraction (XRD) analysis (RIGAKU Ultima IV, Japan) using Cu Kα radiation (λ = 1.5406 Å) with a scan speed of 10° min⁻¹ ranging from 10° to 80°. The morphology of the sample was observed using a field-emission electron scanning microscope (FE-SEM, JEOL, JSM-6500F). Specific surface area was measured by nitrogen physisorption using the Brunauer–Emmett–Teller (BET) method (BELSORP-max, BEL Japan, Inc.). UV-vis diffuse reflectance spectroscopy (DRS) measurements were carried out using a Hitachi UV-3100 UV-vis spectrophotometer equipped with an integrating sphere attachment. The analysis range was from 300 to 650 nm. The band gap energy of the prepared catalysts can be calculated by the following formula:

αhν = A(hν − E_g)^n/2 (1)

where α, ν, E_g and A are absorption coefficient, light frequency, band gap energy, and a constant, respectively. Among them, n is determined by the type of optical transition of a semiconductor (n = 1 for direct transition and n = 4 for indirect transition). The values of n for AgBr and m-WO₃ are 4 and 1, respectively. X-ray photoelectron spectroscopy (XPS) measurements were done on a KRATOA XSAM800 XPS system with Mg Kα source. All the binding energies were referenced to the C 1s peak at 284.8 eV of the surface adventitious carbon.

Evaluation of photocatalytic activities of AgBr/m-WO₃

The photocatalytic properties of the as-obtained AgBr/m-WO₃ samples were tested using a home-made system equipped with a 400 W metal–halogen lamp (IWASKI ELECTRIC Co., Ltd, Japan) and ultraviolet cut-off filter (UV-cut 420), providing a visible-light source (λ ≥ 420 nm). The aqueous solutions of rhB, MO and MB were used as the target pollutants to evaluate the visible-light-driven photocatalytic performance of these samples. The catalyst-free dye solution was analyzed with a HITACHI 200-10 spectrophotometer. All the experiments were conducted in ambient conditions.

Typically, 100 mg of photocatalysts were firstly dispersed in a 100 ml beaker, containing 100 ml of 10 mg L⁻¹ rhB (or MO, MB) aqueous solutions under magnetic stirring in the dark. The distance between the bottom of the lamp and the top of the solution was 20 cm. The above suspensions with dye molecules were kept stirring in the dark for 60 min to reach an adsorption–desorption equilibrium of dye molecules on the photocatalysts. Then, the suspensions with photocatalysts and dye molecules were exposed to the visible light irradiation. At certain time intervals, 3 ml solution were sampled and centrifuged to remove the photocatalyst particles. The top transparent solutions obtained were then transferred to quartz cuvette to measure their absorption spectra in a wavelength range of 400–800 nm. The relative concentrations (C/C₀) of the rhB, MO and MB solutions were determined by the absorbance (A/A₀) at 554 nm, 464 nm and 664 nm, respectively, because of the relationship of C = k_appA. Here, k_app is a constant, A is the absorbance of the dye aqueous solution at time t and A₀ is the absorbance at the beginning of the visible-light irradiation. Similarly, C is the concentration of the dye aqueous solution at time t, and C₀ is the concentration at the beginning of the visible-light irradiation.

For the stability measurements, 100 mg optimum photocatalyst (30 wt% AgBr/m-WO₃) was dispersed in 100 ml of rhB solution (10 mg L⁻¹) and the photoactivity tested as described above. After each photoactivity test, both the separated catalyst particles and the solution were returned to the reduction system. Photocatalysis stability tests were performed on the optimum composition in seven separated runs. Once these tests were completed the photocatalyst was reclaimed, washed, dried overnight at 40 °C, and then analyzed by XRD.

Results and discussion

The stable monoclinic WO₃ (m-WO₃) can have a ReO₃-type structure (corner-sharing arrangement of octahedral). Each W ion is surrounded by six equidistant oxygen ions. An infinite array of corner-sharing WO₆-octahedra is formed like in Fig. 2. The W ions occupy the corners of a primitive unit cell, and O ions bisect the unit cell edges.

Fig. 2 Crystal structure of m-WO₃.

In order to confirm the crystalline structure of prepared AgBr/m-WO₃, powder XRD study was carried out. Fig. 3 shows the XRD patterns of fresh AgBr/m-WO₃ with different silver halide contents. The patterns shows that WO₃ substrates were monoclinic WO₃ (JCPDS: 71-2141) and AgBr was of face-centered cubic structures. As evident form Fig. 3, with increasing silver halide content, the intensities of diffraction peaks of AgBr increased whereas those of substrate decreased simultaneously. The lattice parameters of AgBr/m-WO₃, as listed in Table 1, are in good agreement with those in the literature. The average crystalline sizes of AgBr on the surface of m-WO₃ were calculated by using Scherrer equation, respectively.³¹

where D is taken as crystalline size, K is a constant equals to 0.89, λ is 1.5406 Å, β is the FWHM measured in radians on the 2θ scale, θ is the Bragg angle for the diffraction peaks. The results were also showed in Table 1.

Fig. 3 XRD patterns of AgBr/m-WO₃ samples with different AgBr loadings.

Table 1 Lattice parameters values of AgBr/m-WO₃ photocatalysts calculated from the XRD patters

Photocatalyst	Lattice parameters				Crystallite size of AgBr^a (nm)
	a (Å)	b (Å)	c (Å)	V (Å^3)
a Calculated using Scherrer equation on (2 0 0) diffraction of AgBr.
m-WO3 (JCPDS: 71-2141)	7.297	7.539	7.698	422.88	—
m-WO3	7.293	7.513	7.678	420.58	—
AgBr (5 wt%)/m-WO3	7.278	7.497	7.665	418.98	35.78
AgBr (10 wt%)/m-WO3	7.282	7.503	7.670	419.33	37.91
AgBr (20 wt%)/m-WO3	7.286	7.505	7.674	419.68	40.23
AgBr (30 wt%)/m-WO3	7.291	7.508	7.682	420.06	41.58
AgBr (40 wt%)/m-WO3	7.295	7.515	7.685	421.07	43.03

---

The SEM images and EDX result of AgBr (30 wt%)/m-WO₃ photocatalyst were presented in Fig. 4. From Fig. 4a, the cuboid rods of monoclinic of WO₃ substrates have been prepared by hydrothermal method. The loaded AgBr nanoparticles with average diameter of about 40 nm were uniformly dispersed on the surface of WO₃. The EDX spectrum (Fig. 4b) shows that the composite catalysts are composed W, O, Ag and Br. It would be the direct evidence to prove the AgBr nanoparticles co-dispersed on the surface of WO₃.

Fig. 4 SEM images of (a) AgBr (30 wt%)/m-WO₃ and (b) EDX spectrum of the sample.

Fig. 5 shows the UV-vis diffuse reflectance spectra of pure WO₃, AgBr and AgBr/m-WO₃. It is shown that m-WO₃, AgBr and AgBr/m-WO₃ all exhibited absorption in the visible light region, among which the absorption edge of AgBr/m-WO₃ (471 nm) was lower than that of pure AgBr (479 nm) and m-WO₃ (490 nm). This also suggests that the crystalline size of AgBr supported on the surface of m-WO₃ is smaller than that of pure AgBr in our research.

Fig. 5 DRS spectra of m-WO₃, AgBr and AgBr/m-WO₃ (AgBr 30 wt%).

According to eqn (1), E_g of AgBr was determined from the plot of (αhν)^1/2 versus energy (hν) (Fig. 6a). From the tangent line of the curve, extrapolated to the hν axis intercept, E_g of AgBr was found to be 2.60 eV. Similarly, E_g of m-WO₃ was 2.69 eV (Fig. 6b).

Fig. 6 Plots of (a) (αhν)^1/2 versus energy (hν) for AgBr and (b) (αhν)² versus energy (hν) for m-WO₃.

BET surface area of as synthesized samples compared with that of commercial WO₃ is shown in Table 2. It is revealed form the table that the composite photocatalysts have smaller BET surface area as compared to commercial WO₃. Moreover, BET surface area decreased with increase in AgBr concentrations. Decrease in surface area might be due to the occupation of pores and interstitial lattice positions of m-WO₃ by AgBr nanoparticles and it is the common characteristic of heterogeneous phase of the composite materials. Similar results have been reported for composite photocatalysts.^15,32

Table 2 BET analysis and photodegradation rate constants of the as-synthesized AgBr/m-WO₃ samples

Photocatalyst	BET surface area (m^2 g^−1)	kapp(rhB) (min^−1)	kapp(MO) (min^−1)	kapp(MB) (min^−1)
Commercial WO3	6.179	0.00012	0.00013	0.00012
m-WO3	5.690	0.00036	0.00024	0.00036
AgBr (5 wt%)/m-WO3	5.981	0.00298	0.00220	0.00108
AgBr (10 wt%)/m-WO3	5.682	0.00591	0.00273	0.00188
AgBr (20 wt%)/m-WO3	5.611	0.01287	0.01056	0.00245
AgBr (30 wt%)/m-WO3	5.220	0.02267	0.01637	0.00588
AgBr (40 wt%)/m-WO3	4.704	0.01612	0.01199	0.00361

---

The photocatalystic activities of as-prepared samples were evaluated by the degradation of rhB, MO and MB under visible-light irradiation. Fig. 7a displays the degradation of rhB in the presence with different AgBr contents. Under visible light, WO₃ had weaker photocatalytic activity for rhB though it can absorb visible light. The photocatalytic activity of AgBr/m-WO₃ increased remarkably with increasing AgBr content, but at the higher AgBr level the catalyst photocatalytic activity decreased slightly, suggesting that the optimal AgBr content in AgBr/m-WO₃ existed when the weight ratio 0.3. Moreover, the similar change tendency of degradation efficiency was also displayed in the process of MO (Fig. 7b) and MB (Fig. 7c) degradation as that of rhB degradation. The degradation efficiency of rhB was enhanced and remained unchanged after irradiation for 240 min with increasing AgBr content. According to Langmuir–Hinshelwood (L–H) kinetic model, the k_app of AgBr/m-WO₃ were respectively calculated and displayed in Table 2. The results clearly demonstrate the optimum AgBr content was 0.3 with the maximal degradation rate. Loading of AgBr greater than 30 wt% leads to weaker catalytic activity because the majority of AgBr is not in close contact with the WO₃ and may actively screen light irradiation. Also, the AgBr/m-WO₃ exhibited excellent visible-light-driven photocatalytic efficiencies for the degradation of dyes, which was much higher the TiO₂ (P25). In AgBr/m-WO₃ system, only AgBr is the active component that absorbs visible light to initiate the photocatalytic reaction, so the AgBr content determines the number of electron–hole pairs that participate in the degradation of dyes.^19,33

Fig. 7 The degradation of rhB (a) by different photocatalysts with the same weigh of each visible-light-active component () commercial WO₃; () prepared m-WO₃; () P25; () 5% wt AgBr/m-WO₃; () 10% wt AgBr/m-WO₃; () 20% wt AgBr/m-WO₃; () 30% wt AgBr/m-WO₃; () 40% wt AgBr/m-WO₃. The degradation of MO (b) and MB (c) were under the same conditions as rhB.

As a hybrid semiconductor, the photocatalytic activity of the composite relates directly to the electronic band structures of individual components, which determine the excitation, transportation, and ultimate fate of the photogenerated charge carriers. According to empirical Mulligen electronegativity principle, the conduction and valence band positions of AgBr and WO₃ as the vital factors determined their photo-reactivity. Despite this methodology cannot give the discrete energy levels of individual semiconductors, it is plausible for understanding the transportation of photogenerated charge carriers and is helpful to roughly judge the thermodynamic trends of photoreaction. Because of the complexity of mechanistic energy principle, the charge carrier transportation mechanism based on energy bias law of heterojunctions has been widely accepted to simplify the depiction of photogenerated charge transportation.³⁴

WO₃ has more positive conduction band (CB) bottom and valence band (VB) top than AgBr.^33,34 The possible pathway for the photocatalytic degradation of dyes with AgBr/m-WO₃ photocatalyst was proposed as follows (Fig. 8): during the photocatalytic oxidation process, the AgBr/m-WO₃ system was transformed to be Ag/AgBr/m-WO₃ system. Ag, AgBr and WO₃ can be simultaneously excited to form electron–hole pairs under visible-light irradiation. The photo-excited electrons (e⁻) on CB_AgBr would prefer to flow down to the CB_WO3 crossing the interface, while the photogenerated positive carriers (holes, h⁺) on VB_WO3 would transfer to the VB_AgBr.^35,36 Such transportation of the photogenerated carriers could either extend their transfer path or stabilize the photogenerated holes in the VB_WO3, leading to prolonged lifetime of the charge carriers and successfully hindering the unfavorable recombination of electron–hole (e⁻–h⁺) pairs. Meanwhile, the photogenerated electrons transfer form the CB bottom of AgBr to that of m-WO₃ or are trapped by Ag NPs formed on AgBr particles. At the same time photogenerated holes also move in the opposite direction from the VB top of m-WO₃ to that of AgBr. Besides the intrinsic photoexcitation of AgBr/m-WO₃, the organic dyes may be excited by light irradiation (photo-activation) and the excited dye* will inject electron onto CB of either AgBr or m-WO₃, and consequently dye* gets back to dye.

Fig. 8 Schematic electronic band diagram of the AgBr/m-WO₃ system.

From the photoelectrochemistry point of view, the photogenerated holes on VB_AgBr and VB_WO3 may oxidize organic dyes, while the electrons on the CB_AgBr rather than CB_WO3 are able to reduce soluble O₂ to O₂⁻, an oxidative species that can break down dyes. AgBr are highly dispersed and closely contacted with WO₃ that facilitated the transportation of charge carriers, meanwhile, the holes on VB_WO3 are positive (more oxidative) than those on VB_AgBr, hence such AgBr/m-WO₃ showed higher activity. The VB_WO3 and VB_AgBr situate more positive than the standard redox potentials of OH˙/OH⁻ (1.99 eV) and H₂O₂ (1.77 eV), suggesting their photogenerated holes are far more oxidative than both OH˙ and H₂O₂.^37,38 In addition, the photogenerated hole can not oxidize OH⁻ to form OH˙ because of the more negative (reductive) potential of W⁶⁺/W⁵⁺ (1.10 eV).³⁹ The reduction–oxidation (redox) half reaction W⁶⁺/W⁵⁺ would not happen since the WO₃ is much stable both in air and solution. Therefore, the photodegradation of organic dyes on the AgBr/m-WO₃ can be associated with the photogenerated holes on the valance bands of AgBr and WO₃.

Based on the discussions afore, the mechanism of photocatalytic degradation of organic dyes on the AgBr/m-WO₃ may be proposed, as described in the eqn (3)–(7). We suppose electrons firstly generated, under visible-light irradiation, from the VB of the AgBr/m-WO₃ and photoexcited organic dyes, then hop onto the CBs of the composite photocatalysts. The CB electrons would react with soluble oxygen to form reactive O₂⁻ radicals, which either attack organic dyes or pass to the CB_WO3, correspondingly, the positively charged holes were left on VBs and excited dye (dye*). The photocatalytic degradation reaction would subsequently take place, in which the dye* would be transformed and decomposed into intermediate products and eventually converted to CO₂ and H₂O by both the VB holes and O₂⁻ radicals (eqn (8) and (9)). At the same time, electrons in the VB of AgBr could be excited up to a higher potential edge under visible-light illumination with energy less than 2.95 eV (λ > 420 nm), and these electrons will be trapped by Ag NPs and further react with O₂ adsorbed on the surface of catalyst to generate reactive O₂⁻ that induced the degradation of dyes (eqn (10)).

(1) Photo-excitation and photosensitized process of organic dyes

WO₃–AgBr + hν → WO₃ (e_CB⁻ + h_VB⁺)–AgBr (e_CB⁻ + h_VB⁺) (3)

Organic dye + hν → dye* + e⁻ (4)

e_CB⁻ + O₂ → O₂⁻ (5)

AgBr/WO₃ + dye* → dye⁺˙ + AgBr/WO₃ (e_CB⁻) (6)

AgBr/WO₃ (e_CB⁻) + O₂ → O₂⁻ + AgBr/WO₃ (7)

(2) Simultaneous photocatalytic reaction

Dye⁺˙ + O₂⁻ → intermediate products → CO₂ + H₂O (8)

Dye + h_(AgBr/WO3)⁺ → intermediate products → CO₂ + H₂O (9)

˙O₂⁻ + dye → CO₂ + H₂O (10)

The stability of the AgBr/m-WO₃ is a concern in photocatalysis due to the photosensitive AgBr involves. In order to investigate the stability of the photocatalysts, experiments of visible-light-driven photodegradation of rhB were repeated for certain times on the recycled AgBr/WO₃ catalysts. After several cycling photocatalysis reaction, all the AgBr/m-WO₃ samples partially lost their activity (Fig. 9), while WO₃ remained its initial activity. AgBr (30% wt)/m-WO₃ showed better initial activity and stability than other AgBr/m-WO₃, though it deactivated and the deactivation was accelerated under prolonged visible-light irradiation (Fig. 10). For example, AgBr (30% wt)/m-WO₃ completely decolorized rhB in 240 min in the first photocatalytic cycling test, but took nearly 480 min in the seventh cycling.

Fig. 9 The degradation of rhB by AgBr/m-WO₃ under visible-light irradiation.

Fig. 10 Cycling tests of rhB photodegradation on AgBr (30 wt%)/m-WO₃ samples under visible-light irradiation.

After each cycling test, the particulate AgBr/m-WO₃ photocatalysts were recovered and examined to investigate the evolution of the samples during the photocatalysis stability tests. The crystalline structure of recycled AgBr (30 wt%)/WO₃ samples were characterized by XRD. The diffraction patterns of the fresh and recycled AgBr (30 wt%)/m-WO₃ are compared in Fig. 11. Both the fresh and recycled photocatalysts showed identical Bragg peaks for the m-WO₃ and AgBr species, but new diffraction peaks centered at 38.20° were observed on the recycled catalyst. It can be well defined to cubic Ag. During the photocatalytic oxidation process, AgBr/m-WO₃ system was transformed to be Ag/AgBr/m-WO₃ system, which ascribed to convert some Ag⁺ ions on the surface of AgBr/m-WO₃ to Ag⁰ species under visible light irradiation.⁴⁰ However, after the fifth cycles, the other new diffraction peaks centered at 32.56° were clearly observed. The diffraction peak can be well defined to cubic Ag₂O (Pn3̄ms, ICSD: 03-5540), a p-type oxide semiconductor with narrow bandgap of 1.30 eV.⁴¹ The diffraction intensity of the Ag₂O became stronger with increased cycling times. For example, the XRD signal of Ag₂O is stronger after seventh recycling than after the fifth cycling test. These changes of the recycled AgBr (30 wt%)/m-WO₃ suggest that the AgBr was transformed into Ag₂O along with the gradual deactivation of the photocatalysts under visible-light illumination.

Fig. 11 XRD patterns of the fresh and recycled AgBr (30 wt%)/m-WO₃ sample.

Thus, the AgBr/m-WO₃ sample was examined by X-ray photoelectron spectroscopy, and the results are shown in Fig. 12. The XPS peak for C 1s is due to the adventitious hydrocarbon form the XPS instrument itself. In Fig. 12a, the Ag 3d spectra of AgBr/m-WO₃ consist of two individual peaks at around 373 and 367 eV, which can be attributed to Ag 3d_2/3 and Ag 3d_5/2 binding energies, respectively. To further investigate the chemical state of Ag element, the XPS peaks of Ag element with the composite photocatalyst using for several consecutive photocatalysis experiment with solution of rhB. After the third cycling test, the Ag 3d_5/2 peak is further divided into two different peaks at 367.6 and 368.8 eV, and Ag 3d_3/2 peak is divided into two different peaks at 373.5 and 374.8 eV (Fig. 12b). The peaks at 368.8 and 374.8 eV are attributed to metal Ag⁰ and the peaks at 367.6 and 373.5 eV are attributed to Ag⁺ of AgBr, and those at 368.8 and 374.8 eV to metal Ag⁰ hence confirming the existence of Ag⁰.⁴² From the XPS peak areas, the surface Ag⁰ and Ag⁺ contents are calculated to be 0.33 and 7.8 mol%, respectively. According the XRD results, the phase of Ag₂O was appeared with increasing cycling test. As for the Ag₂O/Ag/AgBr/m-WO₃ nanocomposite, the sample shows similar XPS peaks of Ag element with Ag/AgBr/m-WO₃ phase (Fig. 12c). The calculated contents of the surface Ag⁺ of the corresponding samples are 11.3 mol%, which indicating the formation of Ag₂O in the Ag/AgBr/m-WO₃ composite. The result is good agreement with XRD analysis.

Fig. 12 Ag 3d XPS spectras of (a) the as-prepared AgBr (30 wt%)/m-WO₃; (b) the AgBr (30 wt%)/m-WO₃ used for three consecutive photocatalysis experiments and (c) the AgBr (30 wt%)/m-WO₃ used for seven consecutive photocatalysis experiments with solution of rhB.

The plausible formation process of Ag₂O may be depicted thorough photochemical reaction eqn (10)–(13). During the photocatalysis process, trace amounts of Ag⁺ and Br⁻ would be generated via ionizing AgBr in photocatalysts (eqn (11)), under the visible light irradiation. The Ag⁺ will then react with OH⁻ to generate AgOH, which spontaneously convert to Ag₂O (eqn (12)). The overall reaction is summarized as eqn (13), revealing that AgBr gradually transformed in to Ag₂O and replaced the AgBr active site during photocatalytic reaction. These changes may be accelerated by light irradiation, in a consequence of the photocatalyst deactivation.

AgBr + H₂O + hν → Ag⁺ + Br⁻ + H_aq⁺ + OH_aq⁻ (11)

2Ag⁺ + 2OH⁻ → 2AgOH → Ag₂O + H₂O (12)

In total: 2AgBr + H₂O + hν → Ag₂O + 2H⁺ + 2Br⁻ (13)

On the basis of the photo-degradation mechanism and stability study, the deactivation mechanism of our Ag/AgBr/WO₃ samples can now be proposed. As schematically illustrated in Fig. 13, a small amount of AgBr was transformed into Ag₂O on the surface of WO₃ once visible-light stroke on the Ag/AgBr/WO₃ samples for a certain period, meanwhile, the Ag₂O concentration gradually increased after long-term irradiation. The XRD intensity of Ag₂O was monotonically increased as observed in Ag/AgBr/WO₃, revealing more Ag₂O species were generated, in the multiple recycling photocatalytic tests. Eventually, the Ag₂O would mask the active AgBr sites, resulting in the formation of Ag₂O/Ag/AgBr/WO₃, with reduced photoactivity.

Fig. 13 Schematic evolution of AgBr/WO₃ samples under long-term visible-light irradiation.

The Ag₂O-induced deactivation can be well explained from the photoelectrochemistry point of view as illustrated in Fig. 14. The estimated VB_Ag2O top locates more negative than the potentials of H₂O₂, ˙OH and VBs of WO₃ and AgBr, suggesting VB_Ag2O is a weaker photo-oxidant with limited oxidizing ability for dyes decomposition. On the other hand, the CB_Ag2O potential is more positive than that of AgBr and O²⁻ but more negative than CB_WO3, suggesting the electrons on CB_AgBr would prefer to hop onto CB_Ag2O, which competed with the electron flowing onto CB_WO3 and production of reactive O²⁻. The in situ-generated Ag₂O would also deteriorate the intimate interfaces between AgBr and WO₃. Ag₂O may absorb more light due to its smaller bandgap energy and thus mask the active AgBr sites. Moreover, the narrow bandgap of Ag₂O, implying short electron migration pathway, would inevitably lead to faster recombination of photogenerated charge carriers. Because the WO₃ remains almost unchanged, the photogenerated Ag₂O species should be the dominant factor responsible for the deactivation of the AgBr/WO₃. In general, despite that the crystal of WO₃ in the AgBr/WO₃ remains stable, the transformation of AgBr to Ag₂O would occur during photocatalytic reaction. Such Ag₂O act as stronger recombination centers and would destroy the synergistic interaction between AgBr and WO₃ in photocatalytic reaction, leading to reduce the surface reactivity arisen form AgBr.

Fig. 14 Schematic band alignments of the deactivated Ag₂O/Ag/AgBr/m-WO₃ system.

At the same time, in the Ag₂O/Ag/AgBr/WO₃ system, Ag₂O, Ag, AgBr and WO₃ can be simultaneously excited to form electron–hole pairs under visible-light irradiation. Subsequently the photogenerated electrons transfer from the CB bottom of AgBr to that of WO₃ or are trapped by AgNPs formed on AgBr particles. The photogenerated holes also move in the opposite direction from the VB top of WO₃ to that of AgBr. Probably, electrons in the VB of AgBr could be excited up to a higher potential edge under visible-light illumination with energy less than 2.95 eV (λ ≥ 420 nm), and these electrons will be trapped by AgNPs and further react with O₂ adsorbed on the surface of catalyst to generate reaction O²⁻ that induced the degradation of rhB, MO and MB. At the same time, CB_Ag2O potential is more negative than CB_WO3, suggesting the electrons on CB_AgBr would prefer to hop onto CB_Ag2O, which competed with the electron flowing onto CB_WO3 and production of reactive O₂⁻. It is worth noting that the results are much different from the AgBr/WO₃ systems.

Meanwhile, on the one hand, reactive holes at the VB of AgBr can oxidize Br⁻ ions to Br⁰ atoms that are reactive radical species and degrade rhB, MO and MB;⁴³ on the other hand, the holes generated on AgNPs can also oxidize the rhB, MO and MB directly. In summary, these organic dyes were decomposed by Ag₂O/Ag/AgBr/WO₃ under visible-light irradiation through O²⁻, Br⁰ or direct h⁺ oxidation pathway.

Conclusions

The cuboid rods of m-WO₃ with different amounts of AgBr were successfully prepared and used for photodegradation of organic dyes (rhB, MO and MB) under visible-light irradiation. The initial activity of organic dyes photodegradation on the AgBr/m-WO₃ is closely related to the AgBr loading. The AgBr (30 wt%)/m-WO₃ showed the best initial activity and stability among these photocatalysts. Then enhanced photocatalytic activity is mainly attributed to the band structure. Loading of AgBr greater than 30 wt% leads to weaker catalytic activity because the majority of AgBr is not in close contact with the WO₃ and may actively screen light irradiation.

All the AgBr/m-WO₃ composite photocatalysts are prone to deactivation in the photocatalytic reactions since the AgBr is vulnerable and transformed to Ag₂O under visible-light irradiation. The as-formed Ag₂O reduces the intimate interface of AgBr/m-WO₃ catalysts, masks the light excitation, and serves as recombination center of photogenerated charge carriers, resulting in the deactivation of the catalysts.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (Grant No. 51504101), the Natural Science Foundation of Jiangsu Province (Grant No. BK20150514), the Natural Science Foundation of Jiangsu Provincial Higher Education of China (Grant No. 15KJB430006), the Start-up Foundation of Jiangsu University for Senior Talents (Grant No. 15JDG014).

References

1. H. Tong, S. X. Ouyang, Y. P. Bi, N. Umezawa, M. Oshikiri and J. H. Ye, Adv. Mater., 2012, 24, 229.
2. Y. H. Kim, J. S. Heo, T. H. Kim, S. Park, M. H. Yoon, J. Kim, M. S. Oh, G. R. Yi, Y. Y. Noh and S. K. Park, Nature, 2012, 489, 128.
3. S. Yamazaki, T. Yamate and K. Adachi, Appl. Catal., A, 2013, 454, 30.
4. S. J. Hong, H. Jun and L. S. Lee, Scr. Mater., 2010, 63, 757.
5. J. Zeng, M. Hu, W. D. Wang, H. Q. Chen and Y. X. Qin, Sens. Actuators, B, 2012, 161, 447.
6. G. C. Xi, J. H. Ye, Q. Ma, N. Su, H. Bai and C. Wang, J. Am. Chem. Soc., 2012, 134, 6508.
7. M. Sasidharan, N. Gunawardhana, M. Yoshi and K. Nakashima, Nano Energy, 2012, 1, 503.
8. S. K. Biswas and J. O. Baeg, Int. J. Hydrogen Energy, 2013, 38, 3177.
9. P. M. Kadam, N. L. Tarwal, S. S. Mali, H. P. Deshmukh and P. S. Patil, Electrochim. Acta, 2011, 58, 556.
10. A. K. Bhosale, N. L. Tarwal, P. S. Shinde, P. M. Kadam, R. S. Patil, S. R. Barman and P. S. Patil, Solid State Ionics, 2009, 180, 1324.
11. C. A. Bignozzi, S. Caramor, V. Cristino, R. Argazzi, L. Meda and A. Tacca, Chem. Soc. Rev., 2013, 42, 2228.
12. A. Enesca and A. Duta, Appl. Phys. A: Mater. Sci. Process., 2013, 111, 639.
13. W. Erbs, J. Desilverstro, E. Borgarello and M. Gratzel, J. Phys. Chem., 1984, 88, 4001.
14. R. Abe, K. Sayama and H. Sugihara, J. Phys. Chem. B, 2005, 109, 16052.
15. S. M. Sun, W. Z. Wang, S. Z. Zeng, M. Shang and L. Zhang, J. Hazard. Mater., 2010, 178, 427.
16. J. Cao, B. D. Luo, H. L. Lin and S. F. Chen, J. Mol. Catal. A: Chem., 2011, 344, 138.
17. L. Kong, Z. Jiang, H. H. Lai, R. J. Nicholls, T. C. Xiao, M. O. Jones and P. P. Edwards, J. Catal., 2012, 293, 116.
18. Y. P. Bi, S. X. Ouyang, J. Y. Cao and J. H. Ye, Phys. Chem. Chem. Phys., 2011, 13, 10071.
19. C. H. An, J. Z. Wang, W. Jiang, M. Y. Zhang, X. J. Ming, S. T. Wang and Q. H. Zhang, Nanoscale, 2012, 4, 5646.
20. X. C. Ma, Y. Dai, M. Guo and B. B. Huang, ChemPhysChem, 2012, 13, 2304.
21. W. S. Wang, H. Du, R. X. Wang, T. Wen and A. W. Xu, Nanoscale, 2013, 5, 3315.
22. C. C. Shen, Q. Zhu, Z. W. Zhao, T. Wen, X. K. Wang and A. W. Xu, J. Mater. Chem. A, 2015, 3, 14661.
23. J. Y. Leng, X. J. Xu, N. Lv, H. T. Fan and T. Zhang, J. Colloid Interface Sci., 2011, 356, 54.
24. R. Q. Cabrera, E. R. Latimer, A. Kafizas, C. S. Blackman, C. J. Carmalt and I. P. Parkin, J. Photochem. Photobiol., A, 2012, 239, 60.
25. D. S. Martínez, A. M. Cruz and E. L. Cuéllar, Appl. Catal., A, 2011, 398, 179.
26. P. K. Biswas, N. C. Pramanik, M. K. Mahapatra, D. Ganguli and J. Livage, Mater. Lett., 2003, 157, 4429.
27. Y. Beak and K. Yong, J. Phys. Chem. C, 2007, 111, 1213.
28. K. H. Chang, C. C. Hu, C. M. Huang, Y. L. Liu and C. I. Chang, J. Power Sources, 2011, 196, 2387.
29. X. Q. Gao, X. T. Su, C. Yang, F. Xiao, J. D. Wang, X. D. Cao, S. J. Wang and L. Zhang, Sens. Actuators, B, 2013, 181, 537.
30. J. Y. Li, J. F. Huang, J. P. Wu, L. J. Cao and K. Yanagisawa, Ceram. Int., 2012, 38, 4495.
31. M. Galceran, M. C. Pujol, C. Zaldo, F. Daz and M. Aguil, J. Phys. Chem. C, 2009, 113, 15497.
32. R. Adhikari, G. Gyawali, T. Sekino and S. W. Lee, J. Solid State Chem., 2003, 197, 560.
33. Y. G. Xu, H. Xu, J. Yan, H. M. Li, L. Y. Huang, Q. Zhang, C. J. Huang and H. L. Wan, Phys. Chem. Chem. Phys., 2013, 15, 5821.
34. H. Ehrenreich and D. Turnbll, Harcourt Brace Hovanovich, Publishers, Academic Press, 1991.
35. R. H. Victora, Phys. Rev. B: Condens. Matter Mater. Phys., 1997, 56, 4417.
36. L. Q. Ye, J. Y. Liu, C. Q. Gong, L. H. Tian, T. Y. Peng and L. Zan, ACS Catal., 2012, 2, 1677.
37. J. Tang and J. Ye, Chem. Phys. Lett., 2005, 410, 104.
38. Z. Jiang, F. Yang, G. Yang, L. Kong, M. O. Joes, T. Xiao and P. P. Edwards, J. Photochem. Photobiol., A, 2010, 212, 8.
39. L. Berggren, J. C. Jonsson and G. A. Niklasson, J. Appl. Phys., 2007, 102, 083538.
40. Q. Zhu, W. S. Wang, L. Lin, G. Q. Gao, H. L. Guo, H. Du and A. W. Xu, J. Phys. Chem. C, 2013, 117, 5894.
41. C. Sun, Q. Li, S. Gao, L. Gao and J. K. Shang, J. Am. Ceram. Soc., 2010, 93, 531.
42. P. Wang, B. B. Huang, X. Y. Qin, Z. Y. Zhang, Y. Dai and M. H. Whangbo, Inorg. Chem., 2009, 48, 10697.
43. L. Kuai, B. Geng, X. Chen, Y. Zhao and Y. Luo, Langmuir, 2010, 26, 18723.

---