Fabrication of an olive-like BiVO₄ hierarchical architecture with enhanced visible-light photocatalytic activity

Abstract

The olive-like BiVO₄ hierarchical architecture has been synthesized by using a facile, template-free hydrothermal method. With an average diameter of about 0.5–1 μm, the as-prepared olive-like BiVO₄ is composed of two dimensional nanoplates with a stacked lamellar arrangement. A possible growth mechanism is proposed based on the characterization results of electron microscopy and X-ray powder diffraction, i.e., cooperative crystallization and a self-assembly synergistic process accompanied by Ostwald ripening. The pH value of the system plays a key role in determining the morphology of the final product. The bandgap of the products decreases gradually with the increase of pH values. Ethylene glycol (EG) molecules can effectively inhibit the growth of BiVO₄ crystals, while simultaneously induce preferential orientation growth. Moreover, evaluated by the photocatalytic degradation of methylene blue under visible-light irradiation, the BiVO₄ synthesized with EG exhibits higher photocatalytic activity than the one synthesized without EG, which is ascribed to a larger specific surface area and hierarchical architecture.

---

Introduction

Controllable synthesis of well-defined nano-crystals or micro-crystals with a uniform size and morphology (such as spheres, rods, tubes and plates) have invoked great interest due to the unique optical, electronic and magnetic properties associated with the quantum size effect and potential applications in the energy and the environmental fields.^1–9 Consequently, various nanostructures based on II–VI and III–V semiconductors, metal oxides, and metals have been fabricated through many methods, including precipitation, template direction, electrochemistry deposition, solid state reaction, and hydrothermal or solvothermal method.^6–12 In these methods, the intrinsic growth of crystal plays an important role in determining the final morphology.¹³ Though it is convenient for morphology control to follow the intrinsic crystal growth, the morphology is restricted by the crystal nature and is usually different from initial expectation in some sense. Moreover, utilization of suitable capping agent can provide an alternative means for dynamical modulation of the crystal growth, possibly leading to the formation of different morphology.^14,15

Recently many researchers have synthesized Bi-based oxide semiconductors like BiFeO₃,¹⁶ Bi₂WO₆,¹⁷ BiOI¹⁸ and BiVO₄ (ref. 19 and 20) due to their good photocatalytic activity under visible-light irradiation, which is attributed to the hybridized valence band of O2p and Bi6s orbitals that narrows the bandgap.²¹ Among them, BiVO₄ has been received considerable attention as it can be used in the fields of degradation of organic pollutants and oxygen evolution via water splitting.^22–26 Since the photocatalytic property of BiVO₄ is strongly dependent on its morphology and microstructure,^27,28 many efforts have been devoted to prepare BiVO₄ with different morphology, such as nanoplate,²² nanotube,²⁹ hierarchical microsphere,³⁰ and mesocrystal.³¹ In addition, the photocatalytic property is also closely related to its crystalline phase.³² BiVO₄ has three major crystalline phases, monoclinic scheelite, tetragonal scheelite and tetragonal zircon structure.²⁷ The monoclinic scheelite shows much higher photocatalytic activity under visible-light irradiation than the other two.²⁷ However, its photocatalytic activity is still not high enough for practical applications owing to its low separation efficiency of photogenerated charge carriers. Hence, the monoclinic BiVO₄ with hierarchical architecture and well-defined morphology and uniform dimension is highly desirable as it may exhibit high photocatalytic activity.

Many methods have been used to synthesize the BiVO₄ nanomaterials hitherto, such as hydrothermal method,³³ co-precipitation,³⁴ solid-state reaction,³⁵ aqueous process³⁶ and sonochemical route.³⁷ The crystal phase and morphology of BiVO₄ can be controlled by adjusting the synthesis conditions. One classical method is by using hydrothermal technique via mixing ammonium metavanadium and bismuth nitrate at a given concentration and then controlling the pH value of the solution.³⁸ The hydrothermal method shows great potential due to its simplicity in preparing monoclinic BiVO₄ with good crystallinity and controllable morphology in an environment benign route. The highly crystalline monoclinic BiVO₄ with rod-like and nanofibrous morphology has been synthesized via a hydrothermal process, which showed good photocatalytic activity of oxygen evolution.²⁰ The monoclinic BiVO₄ nanosheets have been prepared via the solvothermal method with sodium dodecylbenzene sulfonate as the morphology-directing agent,³⁹ and the nanoellipsoids have been obtained by using oleic acid.⁴⁰ Both the monoclinic nanosheets and nanoellipsoids have been used for degradation of Rhodamine B under visible-light irradiation. In addition, the monoclinic BiVO₄ single crystalline with different morphology has been synthesized using a tri-block copolymer P123 as hard template via a hydrothermal route.⁴¹

It is noted that most nanostructures were synthesized in the presence of surfactants, which not only increases the cost but also makes it difficult for post-treatment. For this concern, the monoclinic BiVO₄ nanoplates with exposed {001} facet have been prepared without any templates or organic surfactants, which still exhibited remarkable photocatalytic activity.²² The monoclinic BiVO₄ with exposed {040} facet has been made by TiCl₃-assisted hydrothermal process without any templates and showed good photocatalytic activity for water oxidation.²³ In addition, ethylene glycol (EG) is widely used to control the morphology of oxide nanomaterials.^42,43 For instance, the hollow olive-shape BiVO₄ with a monoclinic scheelite structure has been fabricated through sodium bis(2-ethylhexyl)sulfo-succinate-induced aggregation and Ostwald ripening process in a mixed solvent of EG and H₂O.²⁸

Herein, we present the synthesis of olive-like BiVO₄ with relatively high visible-light photocatalytic activity by using a facile, template-free approach. The morphology can be modulated by the EG molecules and the self-assembly of two dimensional (2D) nanoplate into hierarchical architecture can be realized by varying the pH value. Such structure exhibits the characteristics of relatively large surface area and multiple scattering, which can facilitate the photocatalysis as the former is in favor of adsorption of organic molecules and the latter can improve light harvesting.

Experimental

Preparation of nanomaterials

All the chemical reagents used in this work were of analytical grade without further purification. Bismuth nitrate penta-hydrate (Bi(NO₃)₃·5H₂O) was bought from Alfa Aesar Co., Ltd. Ammonium metavanadate (NH₄VO₃) and sodium hydroxide were purchased from Sinopharm Reagent Chemical Co., Ltd. The EG was bought from Shantou Xilong Chemical Factory Co., Ltd. Milli-Q water was used in all experiments.

Typically, 6 mmol of Bi(NO₃)₃·5H₂O was added into 15 mL of a mixed solvent of EG and H₂O (with a volume ratio of 2 : 1) under stirring, which was denoted as solution A. A yellow suspension (solution B) was obtained after vigorous stirring for 20 h of the mixture of 6 mmol of NH₄VO₃ and 30 mL of EG. Then, solution B was added into solution A dropwisely. The obtained solution became transparent orange after stirring for 20 min, for which the pH value was adjusted from 2.05 to 6.0 with 2 M NaOH aqueous solution. Finally this mixed solution was transferred into a Teflon-lined stainless steel autoclave (100 mL) and heated at 180 °C for 24 h. After the autoclave was cooled down naturally to room temperature, the obtained precipitate was separated by centrifugation, washed with absolute ethanol for several times and then dried in vacuum oven at 70 °C for 12 h. For comparison, the BiVO₄ was also synthesized with the same protocol as above except using H₂O instead of EG as solvent (i.e., without EG).

Characterizations

X-ray powder diffraction (XRD) patterns were recorded with a Bruker D8 focus diffractometer with Cu Kα radiation (λ = 0.154184 nm). The average crystallite size was calculated using Scherrer equation (D = kλ/β cos θ) after correcting instrumental broadening. The morphology of the samples was observed by field-emission scanning electron microscopy (FE-SEM, Hitachi, Japan). Transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) were conducted with a Tecnai G2 F20 U-TWIN electron microscope (FEI, USA) using a 200 kV accelerating voltage. The Brunauer–Emmett–Teller (BET) surface area of the samples was analyzed by nitrogen adsorption using a Micromeritics ASAP 2020 nitrogen adsorption apparatus (USA). All samples were degassed at 100 °C prior to nitrogen adsorption measurements. The BET specific surface area was determined by a multipoint BET method using adsorption data in the relative pressure (P/P_o) range of 0.05–0.30. Desorption data were used to determine the pore size distribution via the Barret–Joyner–Halender (BJH) method, given cylindrical pore model. The volume of adsorbed nitrogen at the relative pressure (P/P_o) of 0.97 was used to determine the pore volume and average pore size. The surface charge of BiVO₄ particles in aqueous solution was measured by a zeta potential analyzer (Zeta-sizer Nano ZS, Malvern). The UV-vis diffuse reflectance spectra (DRS) were obtained with a UV-vis-NIR spectrometer (Perkin Elmer Lambda 750) and were converted from reflection to absorption by using the Kubelka–Munk function. The photoluminescence (PL) spectra were measured on a Luminescence Spectrometer (LS-55, Perkin Elmer, USA) at room temperature. The excitation wavelength was 450 nm and the scanning speed was 200 nm min⁻¹. Both the slit width of excitation and emission were 10.0 nm.

Photocatalytic activity test

The photocatalytic degradation of methylene blue (MB) in aqueous solution under visible light (300 W xenon lamp, λ > 400 nm) was used to evaluate the photocatalytic activity of the obtained olive-like BiVO₄ catalysts. Typically, 0.08 g of as-prepared BiVO₄ powder was added into 80 mL of 10 μM MB aqueous solution in a container. The suspension was treated via ultrasonication for 15 min and stirred in the dark for 1 h to reach the adsorption/desorption equilibrium before irradiation. The MB solution containing the photocatalyst was continuously stirred by a magnetic stirrer during the photocatalysis. The MB concentration was monitored by checking the absorbance at 664 nm using the above UV-vis absorption spectrometer.

Results and discussion

Crystal structure

Fig. 1a is the XRD patterns of the BiVO₄ synthesized at different pH values. All the samples exhibit characteristic diffraction peaks at around 15.1, 18.6, 18.9, 28.6, 28.8, 28.9 and 30.5°, which agree well with those of the monoclinic phase BiVO₄ (JCPDS card no. 75-1867, a = 0.51956 nm, b = 0.50935 nm, c = 1.17044 nm, γ = 90.383°). Thus all the obtained BiVO₄ samples are of monoclinic scheelite structure (space group I2/b).^28,40 The cell parameters of the as-prepared products at different pH values are calculated by using the least-square refinement method (Table 1), which are similar to those of the standard monoclinic scheelite BiVO₄. The cell parameter “a” increases linearly with the increase of the pH value, while the parameter “b” and “c” initially increase with the increase of the pH value and then decrease when the pH further increases. This implies that the pH value in the synthesis system can have an impact on the growth of BiVO₄ nanocrystals.

Fig. 1 (a) XRD patterns of the BiVO₄ prepared with EG at different pH values, (b) zoom-in of peaks (112), (004) and (011), (c) correlation of pH value with peak intensity ratio of (011)/(112) and (004)/(112), for which the intensity was normalized with (112) peak as an internal standard.

Table 1 Cell parameters of the obtained monoclinic scheelite BiVO₄

Sample	a (nm)	b (nm)	c (nm)	γ (°)	Volume (nm^3)
PDF 75-1867	0.51956	0.50935	1.17044	90.383	0.30974
pH = 2.05	0.51746	0.51082	1.16806	90.170	0.30875
pH = 3.00	0.51839	0.51153	1.16857	90.265	0.30987
pH = 4.02	0.51922	0.51191	1.16846	90.216	0.31057
pH = 5.01	0.52016	0.51041	1.16800	90.460	0.31009
pH = 6.00	0.52022	0.50889	1.16780	90.354	0.30915

---

The zoom-in patterns are used to observe the changes in intensity of the major diffraction peaks more clearly among the samples synthesized at different pH values (Fig. 1b). As the pH value increases from 2.05 to 6.00, the intensity of the strongest diffraction peak (112) increases initially and then decreases gradually with a max value at pH of 3.00. The relative intensity of the (004) peak increases obviously as the pH value increases, while the intensity of the (011) peak decreases gradually. The peak intensity is normalized by using (112) peak as the internal standard so as to acquire rational comparison. The dependence of the intensity ratio of (011)/(112) and (004)/(112) on the pH value is shown in Fig. 1c. The intensity ratio of (004)/(112) increases gradually when the pH value increase from 2.05 to 4.02 and increases sharply as the pH value further increases to 6.00; while the (011)/(112) peak ratio hardly changes with the change in pH. This indicates that a higher pH value (i.e., less acidity) can facilitate the formation of (004) crystal plane with the experimental conditions used in this work.

The influence of EG molecules present in the synthesis system on the crystal structure is studied by using the sample prepared at a pH of 2.05 as the example. For the sample prepared without EG, although most of the diffraction peaks in the XRD pattern can still be indexed to the monoclinic scheelite BiVO₄ (Fig. 2), the impurities can be also observed in the XRD pattern (denoted as asterisks in Fig. 2). It is found that the full width at half maximum (FWHM) for almost all of the diffraction peaks decreases for the sample prepared without the EG. In addition, it is noted that the intensity of peak (004) is much weaker for the sample prepared with EG, indicating that the preferred growth of the crystal along the normal of (001) plane. The crystal size of BiVO₄ prepared with EG at pH of 2.05 (21.6 nm) is smaller than its counterpart synthesized without EG (31.3 nm). Hence, the presence of EG molecules in the synthesis system can inhibit the growth of BiVO₄ crystals, specifically the development of the (001) facet.

Fig. 2 XRD pattern of the BiVO₄ sample synthesized at pH of 2.05 in the absence of EG.

Energy band structure

Fig. 3a shows the UV/Vis diffuse reflectance spectra (DRS) of BiVO₄ samples synthesized at different pH values. For all of the samples, the strong absorption extends from the UV region to visible light. The steep absorption edge indicates that the absorption is not due to the transition from any impurity levels but to the bandgap transition. The bandgap can be determined with the formula αhν = A(hν − E_g)^n/2, where α, h, ν, A, E_g, and n are the absorption coefficient, Plank's constant, incident light frequency, constant, bandgap energy, and the integer, respectively. The value of integer n is dependent on the characteristics of optical transition in a semiconductor, i.e., direct transition (n = 1) or indirect transition (n = 4). For BiVO₄, it is direct transition and the value of n is thus 1.^28,44,45 The bandgap of BiVO₄ can thus be estimated from a plot of (αhν)² versus the photon energy (hν). The linear intercept to the x axis gives a good approximation of the bandgap, which is 2.57, 2.56, 2.53, 2.49 and 2.46 eV for the as-prepared BiVO₄ samples prepared at pH of 2.05, 3.00, 4 0.02, 5.01 and 6.00, respectively. These values are slightly larger than the reported value 2.4 eV,^46,47 which is possibly due to the nanosize effect. Thus the bandgap of the BiVO₄ samples decreases slightly with the increase of the pH values, while the change is too small to be ascribed to any specific reason. For both the samples prepared at pH of 2.05, in addition, the bandgap for the one prepared with EG is slightly larger than that prepared without EG, which agrees with the trend of the change in the crystal size.

Fig. 3 (a) UV-vis diffuse-reflectance spectra and (b) plots of (αhν)² versus photon energy (hν) of the BiVO₄ samples synthesized at different pH values.

For the monoclinic scheelite BiVO₄, conduction band (CB) is mainly composed of V_3d orbital, and the valence band (VB) is formed by a hybridization of the Bi_6s and O_2p orbitals.²⁰ The CB minimum of a semiconductor at the point of zero charge can be predicted by the formula of E_CB = χ − E⁰ − E_g/2, where χ is the absolute electronegativity of a semiconductor, expressed as the geometric mean of the absolute electronegativity of the constituent atoms (defined as the arithmetic mean of the atomic electron affinity and the first ionization energy); E⁰ is the energy of free electrons on the hydrogen scale (ca. 4.5 eV); E_g is the band gap of the semiconductor.⁴⁸ The value of χ is 6.035 for BiVO₄.⁴⁸ According to the E_g value estimated above, the E_CB of samples synthesized at pH of 2.05, 3.00, 4.02, 5.01 and 6.00 is thus calculated to be 0.25, 0.26, 0.27, 0.29 and 0.31 eV, respectively. Hence, the corresponding E_VB of the samples is 2.82, 2.82, 2.80, 2.78 and 2.77 eV, respectively. So the electronic structure of BiVO₄ changes very little with the varying pH during the synthesis.

Morphology characteristics

The influence of pH value of the precursor on the morphology of the final BiVO₄ products has been studied by SEM (Fig. 4). When the EG is used during the synthesis, the obtained BiVO₄ nanomaterials exhibit hierarchical architecture with a size of 0.5–1 μm, which are dispersed uniformly and are formed by densely-stacked 2D lamellar nanoplates with a size of 50–100 nm. It is noted that the 2D nanoplates are comprised of small crystals with an XRD size of 20–24 nm (Table 2). Such hierarchical architecture is stable and cannot be broken into discrete nanoplates after 10 min of ultra-sonication. Moreover, it appears in different shape with varying pH value. When the BiVO₄ sample is synthesized at pH of 2.05, it appears olive-like shape (Fig. 4a); while it gradually turns into sphere when the pH increases from 2.05 to 4.02 (Fig. 4a–c) and then into cuboid when the pH further increases up to 6.00 (Fig. 4d and e). However, it is noted that the BiVO₄ prepared without EG has an irregular dendritic morphology (Fig. 4f), which is similar to the previous report.²² This indicates the key role of EG present in the synthesis precursor for the formation of the hierarchical architecture. Thus, the BiVO₄ with different hierarchical architecture can be prepared successfully by using the template-free solvothermal method in an EG–H₂O mixture solvent via adjusting the pH value of the precursor.

Fig. 4 Typical SEM images of BiVO₄ samples synthesized with (a–e) and (f) without EG at different pH values, (a) 2.05, (b) 3.00, (c) 4.02, (d) 5.01, (e) 6.00, and (f) 2.05.

Table 2 Effect of pH value on the physical properties of BiVO₄ samples

Sample	Crystalline size (nm)	SBET (cm^2 g^−1)	Pore volume (cm^3 g^−1)	Average pore size (nm)
pH = 2.05	21.6	6.01	0.020	13.52
pH = 3.00	23.9	4.21	0.014	13.71
pH = 4.02	20.7	3.16	0.012	15.22
pH = 5.01	20.2	1.51	0.005	13.74
pH = 6.00	22.7	1.35	0.004	12.72
pH = 2.05 without EG	31.3	1.93	0.003	6.69

---

The characteristics of the hierarchical architecture are further investigated by TEM. Here the olive-like nanomaterials prepared at pH of 2.05 are used as the example, as it exhibits the highest activity for the photodegradation of MB as shown later. It is known from the above SEM images that the length of the long and short axis of the olive-like particles is about 0.5 and 1 μm, respectively (Fig. 4a), which is further confirmed by TEM images (Fig. 5a and b). Moreover, the obtained olive-like BiVO₄ sample has solid interior (Fig. 5b), different from the reported hollow interior of the BiVO₄ prepared using EG.²⁸ From the outer edge of a single particle, it can be seen that the olive-like BiVO₄ is composed of many self-assembled 2D irregular nanoplates with the size of tens of nm (Fig. 5b). The selected area electron diffraction (SAED) pattern implies that the olive-like BiVO₄ is polycrystalline in nature (inset of Fig. 5b). A lattice spacing of 0.312 and 0.293 nm is determined from Fast Fourier Transform pattern in the HRTEM images of the olive-like hierarchical architecture (Fig. 5c and d), corresponding respectively to the interplannar distance of (103) and (004) crystal plane in the monoclinic scheelite BiVO₄.

Fig. 5 (a) and (b) TEM images, and (c) and (d) HRTEM images of BiVO₄ sample synthesized at pH = 2.05. Inset of (b) is the related SAED pattern.

The odontoid edge of the TEM image suggests that the obtained olive-like particle is porous to some degree, which is confirmed by the measurements of N₂ adsorption–desorption isotherms (Fig. 6). According to Fig. 6, all the BiVO₄ samples exhibit isotherm of type IV (Brunauer–Deming–Deming–Teller classification), as indicated by a hysteresis loop at high relative pressure associated with capillary condensation of gases with mesopores (2–50 nm). The hysteresis loop is of type H3, which agrees with the slit-shaped pores resulted from aggregates of nanoparticles or microparticles.^22,49 The porosity properties for all the samples are summarized in Table 2. It is found that the olive-like BiVO₄ sample fabricated with EG at pH of 2.05 has the highest specific surface area (S_BET) and pore volume. When the pH value increases from 2.05 to 6.00, the S_BET and pore volume decrease gradually, indicating that the pH value of the precursor solution has a crucial impact on the microstructures of the products. Moreover, the BiVO₄ sample synthesized at pH of 2.05 with the presence of EG shows a much higher S_BET (6.01 cm² g⁻¹) and pore volume (0.020 cm³ g⁻¹) than those of the sample prepared at pH of 2.05 but without EG (1.93 cm² g⁻¹ and 0.003 cm³ g⁻¹, respectively). In addition, all of the samples prepared with the presence of EG have a larger average pore size (about 13–15 nm) than that prepared without EG (∼6.7 nm). Thus, a low pH value of the precursor solution and the presence of EG are in favour of the formation of BiVO₄ with high S_BET and pore volume.

Fig. 6 Nitrogen adsorption–desorption isotherm curves of the BiVO₄ samples prepared at different pH values.

Formation mechanism of hierarchical architecture

The time-dependent experiments have been carried out so as to investigate the formation mechanism of the hierarchical architecture, i.e., the product has been collected at different stage during the reaction. Here the olive-like BiVO₄ is still used as the example. As shown in Fig. 7, a number of nanoparticles with an average size of about 19.5 nm are observed before the solvothermal treatment (i.e., 0 h), which are amorphous in nature according to the XRD pattern (Fig. 8a). After 0.5 h of reaction, the olive-like BiVO₄ with an average size of 0.5–1 μm are formed (Fig. 7), which are composed of densely-stacked nanoplates and are monoclinic scheelite BiVO₄ in light of the XRD pattern (Fig. 8a). After 1 h of reaction, the products are exclusively olive-like structure particles, while the initial amorphous nanoparticles are depleted completely. With further longer reaction time, the nanoplates are assembled well into olive-like structure (Fig. 7), while the crystalline size determined from XRD patterns using Scherrer equation hardly changes (about 22–23 nm). By using (112) diffraction peak as the internal standard for normalization, it is found that the XRD peak intensity ratio of (004)/(112) increases in the beginning of the reaction, meanwhile the ratio of (011)/(112) decreases (Fig. 8b). After 4 h of reaction, only small changes can be observed for both ratios, while still towards to different direction. This indicates that the growth along [001] direction is promoted in the initial stage of crystal growth and/or the growth along [011] direction is inhibited to some degree. Such preferential growth no longer exists when the reaction time is long enough (such as longer than 4 h). A plausible mechanism can thus be proposed as shown below.

Fig. 7 Typical SEM images of the BiVO₄ samples synthesized at pH of 2.05 with different reaction time.

Fig. 8 (a) XRD patterns of BiVO₄ samples prepared at pH of 2.05 with different reaction time, (b) correlation of reaction time with intensity ratio of (011)/(112) and (004)/(112) derived from the XRD patterns of the corresponding BiVO₄ samples.

The Bi(NO₃)₃ can be hydrolyzed to produce slightly soluble BiONO₃ in the mixed solvent of EG and water. The obtained white suspension gradually becomes clear after stirring for several minutes, implying that the BiONO₃ is dissolved and the formation of BiO⁺(C₂H₆O₂) via coordination between Bi³⁺ ions and EG molecules. The clear solution can become transparent orange under stirring with the addition of NH₄VO₃. As the pH value of the obtained orange solution is adjusted to 2.05–6.00 by using NaOH, the vanadate anions may exist mainly as poly-orthovanadate anions (V₁₀O₂₈⁶⁻ and/or V₃O₉³⁻).⁵⁰ Due to its low solubility, BiVO₄ is generated firstly in the form of amorphous phase via the reaction between the poly-orthovanadate and BiO⁺(C₂H₆O₂) ions under the hydrothermal conditions. Then the amorphous product is dissolved gradually and re-crystallized into the monoclinic scheelite BiVO₄ phase.

Under thermodynamic and kinetic control, usually there are two crystallization pathways, depending on the free energy of activation associated with the nucleation, growth and phase transformation.¹³ One is that the system follows a one-step route to the final mineral phase and another is that the system proceeds by sequential precipitation. Based on the above results, it is suggested that the formation of the olive-like hierarchical architecture is due to cooperative crystallization and self-assembly process, accompanied with Ostwald ripening. In the first stage, plenty of tiny crystalline nuclei are generated before the solvothermal treatment (Fig. 7) and grow into nanoparticles with amorphous phase and an average size < 20 nm (Fig. 8a). As no surfactants or templates are used, these nanoparticles quickly aggregate together so as to minimize the surface energy.⁵¹ As the monoclinic scheelite phase is observed after solvothermal reaction of 0.5 h (Fig. 8a), it is suggested that the formation of crystalline BiVO₄ undergoes dissolution and recrystallization process, i.e., the aggregated particles start to dissolve into the solution and re-crystallize under a temperature of 180 °C. Accordingly, many crystallized particles with an average size of tens of nm are formed. According to the aforementioned XRD results, the growth of (004) crystal plane is markedly inhibited in the presence of EG molecules. Hence, the obtained crystallized particles grow anisotropically along the normal of (001) plane due to the preferential adsorption of EG molecules, resulting in the formation of 2D nanoplates.

As the Ostwald ripening process proceeds, the nanoplates grow until all the nanoparticles are consumed, accompanied by their self-organization into the olive-like structure through the interaction of intermolecular hydrogen bonding and/or van der Waals' force due to the presence pre-adsorbed EG. The different morphology of the BiVO₄ samples synthesized at different pH value is still unclear. Since the pH value has no apparent impact on the average crystalline size of the obtained products (Table 2), considering the formation of vanadate species is strongly dependent on the pH value of the solution, this may be ascribed to the different vanadium species present in the precursor as well as their different concentration,⁵⁰ which may result in different nucleation–dissolution–re-crystallization rate. This is different from the previous report, for which the assembly usually proceeds with stable preformed crystalline building blocks.^52,53 Though the vanadate anions may exist mainly as poly-orthovanadate anions (V₁₀O₂₈⁶⁻ and/or V₃O₉³⁻) when the pH value is in the range of 2.05–6.00,⁵⁰ however, it is difficult to tell one from another when they co-exist in the solution, let alone the respective role it plays in the formation of the nanomaterials. Therefore, further study is required to probe the formation mechanism of the observed hierarchical architecture.

Photocatalytic activity

The MB, a widely used dye, is chosen as the model pollutant. Its characteristic absorption at about 664 nm is used to monitor the photocatalytic degradation process. Since the pH value of zero point of charge (pH_zpc) for the olive-like BiVO₄ is measured to be 3.28 and the pH of the MB aqueous solution (10 μM) is 5.46 at room temperature, the obtained BiVO₄ is negatively charged in the suspension. As the MB is a cationic dye,⁵⁴ it can thus be easily adsorbed on the surface of BiVO₄ photocatalysts. Fig. 9 shows photocatalytic activity of the samples prepared with different pH values. After visible-light irradiation for 1 h, the percentage of MB photo-degradation over BiVO₄ synthesized at pH of 2.05, 3.00, 4.02, 5.01 and 6.00 is 95.7%, 93.6%, 92.4%, 88.8% and 84.1%, respectively. So the photocatalytic activity decreases with the increase of pH value used for the synthesis of the catalysts. In other words, the photocatalytic activity of the BiVO₄ samples decreases with the increase of XRD peak intensity ratio of the (004)/(112). Thus, the exposure of (004) facet has a negative impact on the photocatalytic degradation of MB molecules. Another reason that may account for such phenomenon is that the specific surface area and pore volume decrease with the increase of pH used for the synthesis (Table 2). It is noted that the average pore size, XRD crystalline size, bandgap of the as-prepared BiVO₄ samples changes very little with the pH value used for synthesis. Moreover, the changes in the visible-light absorption for different samples do not agree with the trend for the photocatalytic activity. So the influence of these factors on the photocatalytic activity of the samples prepared at different pH value is ignored in this work, even in the case that they might have little impact on the photocatalysis.

Fig. 9 Photodegradation of MB over BiVO₄ synthesized at different pH value.

The olive-like BiVO₄ is used as the example to study the stability and reusability of the obtained catalysts. It is found that almost the same activity for the photocatalytic degradation of MB molecules is observed after three runs under the same experimental conditions (Fig. 10a). Moreover, the change can hardly be found in the XRD pattern and SEM image after the recycling tests (Fig. 10b). Thus, the obtained hierarchical BiVO₄ samples are quite stable upon visible-light photocatalysis.

Fig. 10 (a) Recycling test for the photocatalytic degradation of MB over BiVO₄ synthesized at pH = 2.05, and (b) XRD patterns of the BiVO₄ (pH = 2.05) before and after recycling test of the photodegradation of MB. Inset of (b) is the SEM image after photocatalysis.

Photodegradation mechanism

It is known that the visible-light photocatalytic degradation of organic pollutants in the presence of a visible-light responsive photocatalyst probably may undergo through two different photochemical pathways, direct semiconductor excitation and indirect dye sensitization. It has been reported that BiVO₄ can be used for visible-light photodegradation of some colorless organic pollutants like phenol and 2,4-dichlorophenol.^48,55 Here it is found that the BiVO₄ can absorb visible light up to 516 nm due to the transition of electrons from the valence band to the conduction band (i.e., the band–band transition) of the BiVO₄ photocatalyst. So the BiVO₄ may be a visible-light active photocatalyst for the photodegradation of MB under direct semiconductor excitation.

The photocatalytic degradation of MB molecules over BiVO₄ photocatalysts via the direct semiconductor excitation mainly include the following three processes (Fig. 11), (1) optical absorption of the BiVO₄ and generation of e⁻ and h⁺, (2) bulk diffusion and surface transfer of e⁻ and h⁺, and (3) surface oxidation reaction mediated by the h⁺ or the derivative active species like ˙O₂⁻. It is noted that the ˙OH radicals is not the main active species initiating the photochemical process over Bi-based catalyst like BiVO₄ because the redox potential of photogenerated holes in the valence band (Bi₂O₄/BiO⁺, +1.59 eV, vs. SHE) is lower than that for the formation of ˙OH radicals (˙OH/OH⁻, +1.99 eV, vs. SHE).⁵⁶ Since the MB molecules themselves can absorb visible light in the range of 550–700 nm, MB would be the major species that absorbs light at the wavelength λ > 516 nm. Hence, in this case the degradation process can be mainly ascribed to the indirect dye sensitization initiated photocatalytic reaction.

Fig. 11 Possible pathway of the photocatalytic degradation of MB over the BiVO₄, (a) direct semiconductor excitation (SE) and (b) indirect dye sensitization (DS). IP-n (n = 1, 2, 3) denotes different intermediate products.

The indirect dye sensitization initiated photodegradation of MB over BiVO₄ photocatalyst mainly involves the following four processes (Fig. 11),⁵⁶ (1) transform of the MB molecules from ground state into excited state (MB*) upon visible light irradiation due to the intra-molecular π–π* transition, (2) immediate injection of the photogenerated electrons in MB* into the conduction band of BiVO₄, leaving behind the MB cation radicals (˙MB⁺), (3) capture of the photogenerated electrons in the conduction band of BiVO₄ by the dissolved O₂, giving rise to active species like ˙O₂⁻, and (4) the degradation of MB molecules by subsequent reactions between MB cation radicals and the active oxygen species, as the organic pollutant usually cannot be degraded from the starting material directly into CO₂ and water. Here the IP-1, IP-2 and IP-3 are not real products, while they represent the intermediate states via different pathways during the photocatalytic degradation of MB. This is the so-called self-degradation of organic dye upon irradiation. Obviously, the holes left behind do not participate in the degradation reaction in the indirect dye sensitization initiated process. Compared with the results shown in Fig. 9, the efficiency is much lower when a light at about 550 nm is used for the photodegradation of MB (Fig. 12). This means that the efficiency for the indirect dye sensitization process is quite low, possibly due to the slow interfacial charge transfer between the excited MB molecules and the conduction band of BiVO₄. Although both the direct semiconductor excitation and indirect dye sensitization may be present simultaneously in this work, therefore, the direct semiconductor excitation is the predominant process that accounts for the degradation of MB molecules. It is noted that, besides the above blank experiments, direct evidence should be discovered in the on-going study.

Fig. 12 Photocatalytic degradation of MB molecules with and without the catalysts under a 300 W xenon lamp with combination of a VISREF filter and a band pass 550 nm filter.

Conclusions

The monoclinic scheelite BiVO₄ with olive-like hierarchical architecture has been successfully synthesized with the assistance of EG via the hydrothermal process. The olive-like BiVO₄ particles are composed of 2D lamellar nanoplates. By varying the pH value during the synthesis, the BiVO₄ with different morphology can be obtained, in which the EG plays an important role. The presence of EG in the synthesis system can inhibit the crystallization of the BiVO₄ and the preferential orientation growth. The formation of the hierarchical BiVO₄ architecture follows a cooperative crystallization and self-assembly process, accompanied by Ostwald ripening. The obtained olive-like BiVO₄ shows relatively high visible-light activity for MB photodegradation mainly due to the relatively large specific surface area that can facilitate the adsorption of MB molecules and the multiple scattering caused by the hierarchical architecture that can improve the light harvesting, which obeys mainly the mechanism of direct semiconductor excitation. This work gives insight into the simple hydrothermal synthesis of hierarchical architecture and illustrates the feasibility for future potential applications.

Acknowledgements

This work was supported by the Ministry of Science and Technology of China (2015DFG62610) and the National Natural Science Foundation of China (11404074).

Notes and references

1. H. Cölfen and M. Antonietti, Angew. Chem., Int. Ed., 2005, 44, 5576–5591.
2. H. Goesmann and C. Feldmann, Angew. Chem., Int. Ed., 2010, 49, 1362–1395.
3. Z. B. Zhuang, Q. Peng and Y. D. Li, Chem. Soc. Rev., 2011, 40, 5492–5513.
4. Y. D. Liu, J. Goebl and Y. D. Yin, Chem. Soc. Rev., 2013, 42, 2610–2653.
5. Q. F. Zhang, E. Uchaker, S. L. Candelaria and G. Z. Cao, Chem. Soc. Rev., 2013, 42, 3127–3171.
6. M. H. Huang and P. H. Lin, Adv. Funct. Mater., 2012, 22, 14–24.
7. G. V. Freymann, V. Kitaev, B. V. Lotsch and G. A. Ozin, Chem. Soc. Rev., 2013, 42, 2528–2554.
8. H. J. Chen, L. Shao, Q. Li and J. F. Wang, Chem. Soc. Rev., 2013, 42, 2679–2724.
9. W. D. Shi, S. Y. Song and H. J. Zhang, Chem. Soc. Rev., 2013, 42, 5714–5743.
10. C. Bae, H. Yoo, S. Kim, K. Lee, J. Kim, M. M. Sung and H. Shin, Chem. Mater., 2008, 20, 756–767.
11. G. Loget, J. Roche, E. Gianessi, L. Bouffier and A. Kuhn, J. Am. Chem. Soc., 2012, 134, 20033–20036.
12. T. Shinagawa, Y. Ida, K. Mizuno, S. Watase, M. Watanabe, M. Inaba, A. Tasaka and M. Izaki, Cryst. Growth Des., 2013, 13, 52–58.
13. H. Cölfen and S. Mann, Angew. Chem., Int. Ed., 2003, 42, 2350–2365.
14. G. Liu, J. C. Yu, G. Q. Lu and H. M. Cheng, Chem. Commun., 2011, 47, 6763–6783.
15. D. K. Ma, S. M. Huang, Y. Y. Yu, Y. F. Xu and Y. Q. Dong, J. Phys. Chem. C, 2009, 113, 8136–8142.
16. S. Li, Y. H. Lin, B. P. Zhang, Y. Wang and C. W. Nan, J. Phys. Chem. C, 2010, 114, 2903–2908.
17. L. W. Zhang, Y. J. Wang, H. Y. Cheng, W. Q. Yao and Y. F. Zhang, Adv. Mater., 2009, 21, 1286–1290.
18. Q. C. Liu, D. K. Ma, Y. Y. Hu, Y. W. Zeng and S. M. Huang, ACS Appl. Mater. Interfaces, 2013, 5, 11927–11934.
19. J. Q. Yu and A. Kudo, Adv. Funct. Mater., 2006, 16, 2163–2169.
20. P. Cai, S. M. Zhou, D. K. Ma, S. N. Liu, W. Chen and S. M. Huang, Nano-Micro Lett., 2015, 7, 183–193.
21. A. Walsh, Y. F. Yan, M. N. Huda, M. M. Al-Jassim and S. H. Wei, Chem. Mater., 2009, 21, 547–551.
22. G. C. Xi and J. H. Ye, Chem. Commun., 2010, 46, 1893–1895.
23. D. E. Wang, H. F. Jiang, X. Zong, Q. Xu, Y. Ma, G. L. Li and C. Li, Chem.–Eur. J., 2011, 17, 1275–1282.
24. L. Ren, L. Ma, L. Jin, J.-B. Wang, M. Qiu and Y. Yu, Nanotechnology, 2009, 20, 405602–405610.
25. M. Ge, L. Liu, W. Chen and Z. Zhou, CrystEngComm, 2012, 14, 1038–1044.
26. A. Iwase and A. Kudo, J. Mater. Chem., 2010, 20, 7536–7542.
27. S. Tokunaga, H. Kato and A. Kudo, Chem. Mater., 2001, 13, 4624–4628.
28. M. L. Guan, D. K. Ma, S. W. Hu, Y. J. Chen and S. M. Huang, Inorg. Chem., 2011, 50, 800–805.
29. L. Ren, L. Jin, J. B. Wang, F. Yang, M. Q. Qiu and Y. Yu, Nanotechnology, 2009, 20, 115603–115611.
30. D. K. Ma, S. Wang, P. Cai, J. L. Jiang, D. P. Yang and S. M. Huang, Chem. Lett., 2009, 38, 962–963.
31. L. Zhou, W. Z. Wang and H. L. Xu, Cryst. Growth Des., 2008, 8, 728–733.
32. X. Zhang, Z. Ai, F. Jia, L. Zhang, X. Fan and Z. Zou, Mater. Chem. Phys., 2007, 103, 162–167.
33. S. Kohtani, S. Makino, A. Kudo, K. Tokumura, Y. Ishigaki, T. Matsunaga, O. Nikaido, K. Hayakawa and R. Nakagaki, Chem. Lett., 2002, 7, 660–661.
34. Y. Zhao, Y. Xie, X. Zhu, S. Yan and S. Wang, Chem.–Eur. J., 2008, 14, 1601–1606.
35. A. R. Lim, S. H. Choh and M. S. Jang, J. Phys.: Condens. Matter, 1995, 7, 7309–7323.
36. Y. Zhou, K. Vuille, A. Heel, B. Probst, R. Kontic and G. R. Patzke, Appl. Catal., A, 2010, 375, 140–148.
37. A. Iwase, H. Kato and A. Kudo, J. Sol. Energy Eng., 2010, 132, 021106–021110.
38. R. S. Roth and J. L. Waring, Am. Mineral., 1963, 48, 1348–1356.
39. L. Zhang, D. R. Chen and X. L. Jiao, J. Phys. Chem. B, 2006, 110, 2668–2673.
40. Y. F. Sun, C. Z. Wu, R. Long, Y. Cui, S. D. Zhang and Y. Xie, Chem. Commun., 2009, 45, 4542–4544.
41. X. Meng, L. Zhang, H. X. Dai, Z. X. Zhao, R. Z. Zhang and Y. X. Liu, Mater. Chem. Phys., 2011, 125, 59–65.
42. R. W. J. Scott, N. Coombs and G. A. Ozin, J. Mater. Chem., 2003, 13, 969–974.
43. X. C. Jiang, Y. L. Wang, T. Herricks and Y. N. Xia, J. Mater. Chem., 2004, 14, 695–703.
44. L. Zhou, W. Wang, S. Liu, L. Zhang, H. Xu and W. Zhu, J. Mol. Catal. A: Chem., 2006, 252, 120–124.
45. L. Zhou, W. Z. Zhong, H. L. Xu, S. M. Sun and M. Shang, Chem.–Eur. J., 2009, 15, 1776–1782.
46. K. Sayama, A. Nomura, Z. Zou, R. Abe, Y. Abe and H. Arakawa, Chem. Commun., 2003, 39, 2908–2909.
47. A. Kudo, K. Omori and H. Kato, J. Am. Chem. Soc., 1999, 121, 11459–11467.
48. M. C. Long, W. M. Cai, J. Cai, B. X. Zhou, X. Y. Chai and Y. H. Wu, J. Phys. Chem. B, 2006, 110, 20211–20216.
49. H. Jiang, H. Dai, X. Meng, K. Ji, L. Zhang and J. Deng, Appl. Catal., B, 2011, 105, 326–334.
50. L. Sun, Y. Zhang, J. Zhang, C. Yan, C. Liao and Y. Lu, Solid State Commun., 2002, 124, 35–38.
51. P. W. Voorhees, J. Stat. Phys., 1985, 38, 231–252.
52. L. S. Zhang, W. Z. Wang, L. Zhou and H. L. Xu, Small, 2007, 3, 1618–1625.
53. Y. Li, J. Liu, X. Huang and G. Li, Cryst. Growth Des., 2007, 7, 1350–1355.
54. L. G. Devi, B. N. Murthy and S. G. Kumar, Chemosphere, 2009, 76, 1163–1166.
55. B. Zhou, X. Zhao, H. Liu, J. Qu and C. P. Huang, Sep. Purif. Technol., 2011, 77, 275–282.
56. H. Fu, C. Pan, W. Yao and Y. Zhu, J. Phys. Chem. B, 2005, 109, 22432–22439.

---