Open Research Online A fair comparison between bismuth catalysts for application in photodegradation under visible and solar light

- Three different bismuth catalysts (Bi 2 WO 6 , BiOI and BiVO 4 ) were synthesized using solvo-and hydrothermal methods. Different reaction times, calcination and the addition of poly (vinyl pyrrolidone) during synthesis were tested to investigate the effect of these variables on the catalysts’ morphology and photocatalytic activity. The photocatalytic activity was evaluated using the degradation of rhodamine B dye under both visible light and natural solar radiation. The Bi 2 WO 6 samples presented good crystallinity and morphological similarities, despite having undergone different treatments. The BiOI and Bi 2 WO 6 catalysts presented a spherical shape, and no morphological difference was observed as a result of the addition of PVP. The BiVO 4 sample presented a parallelepiped shape. BiOI containing PVP and ethylene glycol was the catalyst that presented the highest activity, while BiVO 4 presented the lowest. In experiments using scavengers, photogenerated holes demonstrated a key role in dye degradation.


INTRODUCTION
Heterogeneous photocatalysis, an Advanced Oxidation Process (AOP) considered to be a sustainable and low operating cost technology, is a suitable wastewater treatment process for effluents containing refractory organic compounds. However, until now, the most common catalysts used in this process are mainly active under UV light, which makes up only about 4% of the sunlight spectrum (Sivakumar et al., 2014;Issarapanacheewin et al., 2016;Meng and Zhang, 2016). Therefore, in order to maximize the effect of solar radiation, it is necessary to develop catalysts that are active under visible light irradiation.
In this context, the use of bismuth-based catalysts has attracted great attention because the hybridization of O -2p and Bi -6s levels generates semiconductors with smaller band-gap and a more disperse valence band (VB). This largely disperse VB favors the mobility of the photo holes and is useful for oxidation reactions (Kudo and Hijii, 1999). Furthermore, bismuth is considered to be a non-toxic and low-cost material (Bonné et al., 2017;Kim et al., 2017).
The photocatalytic activity of these catalysts has been widely studied and improved through various methods. Among them are hydro/solvothermal Lin et al., 2014;Wu, D. et al., 2016), sol-gel (Wu et al., 2010;Zhang et al., 2010a;Zhang et al., 2010b), ultrasound (Zhou et al., 2006;Zhou et al., 2007;Dong et al., 2014) and coprecipitation (Alfaro and Martínez-De La Cruz, 2010;Ravidhas et al., 2015). The hydro/solvothermal method is the most widely used, primarily due to its low cost, simplicity, low temperatures and short reaction time. Furthermore, factors that affect the catalyst's activity such as morphology, crystal structure and band gap can easily be controlled by this method.
Although there are several published studies, to the best of our knowledge a fair comparison of photocatalytic activity between different types of bismuth catalysts under the same experimental conditions has not yet been done.
Therefore, in this work, hydro/solvothermal methods were used in the synthesis of three different bismuth catalysts: Bi 2 WO 6 , BiOI and BiVO 4 . The relation between the synthesis conditions and the photocatalytic activity of the material obtained was investigated using rhodamine B (RhB) dye. The evaluated variables were: synthesis time, calcining and the addition of polyvinylpyrrolidone (PVP) to Bi 2 WO 6 and BiOI.

Synthesis of Bi 2 WO 6
The typical synthesis was based on the work of Zhang et al. (2007a), in which 0.98 g of Bi(NO 3 ) 3 .5H 2 O was dissolved in 30 mL of 0.4 M nitric acid solution. The mixture was then stirred for several minutes at 40 °C. A white precipitate was formed when 20 mL of a 0.05 M Na 2 WO 4 (with or without 0.15 g PVP K30) solution was added dropwise into the mixture. After being stirred for 24 hours, the suspension was added to a 70 mL Teflon -lined autoclave and maintained at 160 °C for 8, 16, 24 and 48 hours. The resulting powders were collected, washed with deionized water and dried in an oven at 80 °C. Calcination, when performed, occurred at 500 °C for 2 hours. Calcined samples, containing PVP, were submitted to 16 h in the autoclave.

Synthesis of BiOI
In the synthesis of the BiOI samples, based on Hu et al. (2014), 0.004 mol of Bi(NO 3 ) 3 .5H 2 O was poured into 40 mL of ethylene glycol (EG). Next, another 40 mL of the corresponding solvent solution, containing 0.004 mol KI (solution A), was added. The mixture was further stirred for 30 minutes, placed in a 70 mL Teflon -lined autoclave and kept at 160 °C for 16 hours. In the synthesis with PVP, 0.15 g PVP K30 was added to solution A. The synthesized precipitates were washed with ethanol, ethanol/water and water (separately and in this order) and dried in an oven at 60 °C for 12 hours.

Synthesis of BiVO 4
The BiVO 4 samples were synthesized according to Ma et al. (2015). Solution A contained 6.0 mmol of Bi(NO 3 ) 3 .5H 2 O dissolved in 20 mL of 4 M HNO 3 . Solution B consisted of 6.0 mmol of NaVO 3 and 3.0 mmol of EDTA dissolved in 20 mL of 4 M NaOH. Both were stirred until total dissolution was achieved. Then, solution B was added dropwise into solution A and the pH was adjusted to 7, using a NaOH solution. After further stirring for 30 minutes, the mixture was transferred into a 70 mL Teflon -lined autoclave and maintained at 160 °C for 4 hours. The sample was washed in the same way as the BiOI samples and dried in an oven at 100 °C for 3 hours.

Characterization
The catalysts were characterized by scanning electron microscopy (SEM, JEOL JSM 6060), operating at 10 kV. The X-ray diffraction data were detected via a D500 Siemens diffractometer, using CuKα (λ = 0,154056 nm) radiation. The operating conditions were controlled at 40 kV and 1.75 mA. The UV-vis diffuse reflectance spectra of the samples were recorded with an UV-vis spectrophotometer (Cary 5000 Scan Spectrophotometers, Varian). Nitrogen adsorption-desorption measurements were conducted at 77 K on a Tristar II Krypton 3020 Micrometrics ® . The surface area and the volume were calculated based on Brunauer-Emmett-Teller (BET) and BJH (Barret, Joyner and Halenda) analyses.

Photocatalytic experiments
The photocatalytic activity of the bismuth catalysts was evaluated by studying the degradation of rhodamine B dye (RhB) in an aqueous solution (Figure 1), under 600 W m -2 visible light irradiation, using a LED lamp (Stellatech 13 W) as a light source or under natural solar light (Porto Alegre, Brazil, 30° 01' S and 51° 13' W). In each experiment, 50 mg of photocatalyst were added to a 50 mL RhB solution (25 or 50 mg L -1 ). The experiments were divided into two steps: 1 hour in the dark (30 min in ultrasound and 30 min in magnetic stirring) to ensure the adsorption/ desorption equilibrium and 90 minutes of reaction. At given time intervals, 1 mL suspensions were sampled and centrifuged to remove the photocatalyst powders. The RhB concentration was analyzed through a UVvis spectrophotometer (λ max = 553 nm). In order to determine the role that the •OH radical and the h + hole play in the reactions, isopropanol and EDTA were used as scavengers. All experiments were carried out in duplicate or triplicate, when necessary, and average values used as results.

Catalysts Characterizations Morphology and crystalline phase
The SEM images for the Bi 2 WO 6 catalysts, Figure  2(a-e), show that all samples have a uniform spherical shape with an approximate average diameter between 6 and 8 μm. Different treatments had little change on the samples' morphology and on the average size of the formed spheres. This fact reveals that there is not necessarily a direct relation between synthesis time and an increase in the microsphere organization, leading to flower-like structures, as reported in the literature (Yu et al., 2005;Zhang et al., 2007;. Furthermore, adding PVP did not cause any significant alteration in the samples' morphology ( Figure 1f).
The images of the BiOI samples (Figure 2g-h), in turn, showed that microspheres with an approximate average diameter between 2 and 4 μm are obtained in  the presence of EG as a solvent. The addition of PVP to the BiOI samples -similarly to what occurred with the Bi 2 WO 6 samples -produced no significant effect on the size and morphology of the microspheres.
In the case of the BiVO 4 , the sample showed a three-dimensional structure similar to a parallelepiped (Figure 2i). These parallelepipeds had a roughened surface and an average length between 10 and 12 μm and a width of 5 μm, approximately.
The catalysts' phase and composition were determined by X-Ray diffraction, which demonstrated that all samples presented good crystallinity ( Figure  3). Furthermore, it was observed that the main diffraction peaks of the Bi 2 WO 6 -16h ( Figure 3a Table 1 provides the specific area, volume, average pore diameter and band gap energy values for the catalysts synthesized in this work. With regards to the Bi 2 WO 6 samples, significant area and volume differences were noticed only when the catalyst was subjected to calcination. Furthermore, the PVP addition increased the sample specific surface area by approximately 40%.
BiOI revealed a mesoporous concentration between 2 and 12 nm for the BiOI-EG-PVP sample and a wider distribution, between 5 and 30 nm, for the BiOI -EG. The smaller size of the pores contributed to a larger area for the BiOI-EG-PVP.
Overall, it was observed that the samples with spherical shapes (Bi 2 WO 6 and BiOI) had larger surface areas than the parallelepiped shape. For photocatalysis, the catalyst surface area plays an important role, since, in most cases, the larger the illuminated area, the greater the observed number of active sites and degradation.
The optical properties of the samples are shown in Figure 4 and the calculated results are presented   in Table 1. It was observed that all samples can be activated by visible light (λ ≥ 400 nm).

Photocatalytic performance (a) LED lamp
The degradation kinetics of rhodamine B were investigated using the synthesized catalysts, and the experimental results set to a first order kinetics. The degradation after 90 minutes and the (apparent kinetic constant -min -1 ) values calculated are shown in Table 1.
Preliminary tests showed that direct dye photolysis (degradation only caused by radiation) is about 18%. When comparing the Bi 2 WO 6 -8 and 16 hour samples, it was observed that increasing the sample's time in the autoclave leads to an increase in dye degradation from 46% to 63%. Autoclave times greater than 16 hours resulted in reduced rhB degradation, but still higher than the 8 hour sample. The difference in activity, however, cannot yet be explained and further analyses are necessary in order to clarify these results. For example, in the work of , after a photoluminescence analysis, the authors noticed that catalysts synthesized at different times presented distinct lifetimes of photogenerated electron-hole pairs, which explained the different photocatalytic performance.
When adding PVP to the synthesis of Bi 2 WO 6 , it is noted that the surfactant does not influence rhB degradation. This finding contradicts reports in the literature in which the PVP addition promoted the formation of more organized structures with greater number of active sites and, thus, with more activity (Li et al., 2007;Wu. et al., 2007;Dai et al., 2010). Still, it is clear that calcination affects the catalyst performance by greatly reducing the specific area.
For the BiOI catalysts, no significant morphological change was noted. However, the presence of PVP clearly increases the BiOI photocatalytic activity. The reason, as seen, is because the presence of surfactant increases the surface area, thus increasing the amount of active sites available for the reaction.
The BiVO 4 catalyst showed inferior photocatalytic performance when compared to photolysis alone. Therefore, the presence of the catalyst actually has a negative influence on the dye degradation. This result can be attributed to the lower surface area of this catalyst and the presence of the tetragonal crystalline phase. It has been demonstrated that only the monoclinic phase has good photocatalytic activity (Yang et al., 2009).
In general, there is a linear correlation between the illuminated area and the photocatalytic activity, as can be seen in Figure 5a. The figure compares the results of this study with literature data, confirming the importance of obtaining catalysts with larger areas. However, when comparing the band gap values and the total reduction achieved, no correlation was observed ( Figure 5b). So, smaller band gap energy does not necessarily imply greater photocatalytic activity. Of course, the band gap determines the wavelength that will activate the catalyst, but once activated, other characteristics will determine the catalyst efficiency in the degradation of certain chemical species.
During the experiments, it was observed that the higher the rhodamine B adsorption on the surface, the greater the activity (Table 1). Those results suggest that the photogenerated holes have a key role in the degradation process. In order to verify this hypothesis, experiments were performed in the presence of isopropanol (IP) (Gamage Mcevoy et al., 2014) and EDTA (Huo et al., 2012), which act as •OH (hydroxyl radical) and h + (hole) scavengers, respectively. The concentrations of IP and EDTA in the reaction system were 0.01 M. For the purposes of these tests, the samples with better photocatalytic activity in earlier experiments were chosen. The results are presented in Figure 6. The presence of IP did not result in effective inhibition of the degradation of rhodamine B, showing that the hydroxyl radical does not play an important role in dye degradation. On the other hand, the addition of EDTA clearly hampered the catalyst performance by suppressing the holes, indicating that h + plays a key role in the rhodamine B degradation mechanism.

Figure 5.
Comparison between the surface area (a) and band gap (b) and photodegradation for several bismuth catalysts (Dai et al., 2010;Liu, Z. et al., 2012;Bi et al., 2013;Xiao et al., 2014) and this work. Figure 6. Reaction rate constants for photocatalytic degradation of rhB on BiOI-EG-PVP and Bi 2 WO 6 -16h with and without scavengers.
(b) Natural solar light In order to measure the behavior of the synthesized catalysts in the presence of solar radiation, experiments were carried out using the samples that obtained the best results with a LED lamp (Table 2). However, as the experiments were performed on different days, natural interferences did not provide the same irradiation conditions for all the tests. Therefore, a comparison of the different catalysts was not possible. Comparing light sources, however, shows that both catalysts were more efficient in the presence of solar radiation, reaching 94% and 98% rhodamine B degradation for BiOI and Bi 2 WO 6 , respectively. This result is attributed to the higher radiation incidence on the catalysts, since the sun irradiates in a more energetic range than the LED lamp.
In addition, as can also be observed in Table 2, the tungstate catalyst was exposed to higher average radiations, both UV and visible, which explains its greater increase in photocatalytic activity when compared to the oxyiodide catalyst.

CONCLUSION
The results showed that different bismuth catalysts, active in visible light and with well-defined morphologies, were easily synthesized by a hydro/ solvothermal method. For Bi 2 WO 6 catalysts, the synthesis time and the addition of polyvinylpyrrolidone did not cause morphological differences in the samples, and calcination at 500 °C produced samples with smaller surface areas. For BiOI, microspheres with a larger surface area were obtained and the addition of PVP once again did not cause significant morphological difference. It did, however, promote greater activity. The BiVO 4 catalyst showed no satisfactory photocatalytic activity. Tests carried out in the presence of natural solar radiation showed that the bismuth catalysts are capable of degrading up to 98% of rhodamine dye. For the photocatalytic process using sunlight, two of the catalysts studied in this work proved to be viable alternatives. However, there is still a long way to go before this material is commercially available.