Solid photocatalysts typically used for degradation processes.
Three alternative approaches for the development of heterogeneous photocatalysts are comparatively evaluated, namely (i) the use of molecular imprinting concept for the development of heterogeneous catalysts employing rhodamine B as template and sol–gel as synthesis route; (ii) the impregnation of TiCl4 on mixed nano- and micro-metric silicas, followed by calcination; (iii) the use of industrial and academic chemical residues as source of potential photocatalyst species impregnated on supports. All tests were carried on with rhodamine B as target molecule. For comparative reasons, photocatalytic tests were carried out with commercial titania (P25). The solids were characterized by nitrogen porosimetry, small-angle X-ray scattering (SAXS), zeta potential (ZP), diffuse reflectance spectroscopy in the ultraviolet region (DRS-UV), diffuse reflectance infrared Fourier transmission spectroscopy (DRIFTS), and Rutherford backscattering spectrometry (RBS). The supported catalysts resulting from silica nanoparticles and residue of the petrochemical industry achieved higher percentage of the dye degradation under ultraviolet (68.0 and 66.8%, respectively) radiation. The industrial waste reached the highest photocatalytic activity under visible (61%) radiation, while the commercial P25 achieved 82.0and 12.3% for ultraviolet and visible radiation, respectively. The textural and structural characteristics of the supported catalyst prepared with fumed silica and petrochemical waste (SiPe), namely the low-energy bandgap (1.8 eV), large surface area (280 m2 g−1), high pore volume (1.9 cm3 g−1), and high zeta potential value (−36.4 mV), may have been responsible for their high activity.
- supported photocatalyst
- Rhodamine B
- molecular imprinting
Advanced oxidation processes (AOPs) are environmentally friendly technologies for the removal of organic pollutants in water and waste water. AOPs generate powerful oxidizing species, such as hydroxyl radicals (•OH, E0 = 2.80 V) which can fragment, destroy, and degrade contaminants into small molecules (1, 2). The oxidation processes (i.e., mineralization) involve the production of CO2, H2O, and, eventually, inorganic ions as end products (3). The effectiveness of an AOP is proportional to its ability to generate hydroxyl radicals. This is dependent on: (i) the chemical and physical mechanisms of this technology; (ii) the nature and concentration of target contaminant; (iii) background water quality; (iv) reactor contact time; and (v) reactor configuration (3, 4).
Among the AOPs, heterogeneous photocatalysis is one of the most promising to degrade emerging contaminants and it is based on the use of bandgap semiconductors metal oxide as catalyst (3, 4). When these materials are being exposed to light with energy equal or superior than their bandgap they tend to generate electron-hole pairs which increase the number of free carriers and hence the conductivity. The ideal characteristics of a catalyst are: low raw material cost, relatively simple processing, useful temperature range, very high reaction, photochemical stability, and ability to adsorb reactants on the particle surface (3). Moreover, this method shows clear advantages, such as low cost, low toxicity, and chemical stability (5). However, once it is affected by photogenerated holes and hydroxyl radicals with a strong oxidation potential, a serious shortcoming in heterogeneous photocatalytic oxidation is the low selectivity for the most hazardous contaminants (6).
The lack of selectivity is a disadvantage, as mixtures derived from effluent streams may contain hazardous contaminants and low toxicity contaminants. In many cases, the former is present in lower concentrations, and the latter is the majority (6), but it is desirable to preferentially degrade the most toxic materials. Other drawbacks include the necessity of improving the contact between pollutant molecules and the catalyst. Regarding to this issue, the low surface area commercial catalysts, which in addition demand UV-light for the oxidative process, and UV-light in natural sunlight represents only 5–8% of the solar spectrum (7). Considering these drawbacks, alternative approaches in development of heterogeneous titania-based photocatalyst are an important issue in the topic of photocatalysis. Table 1 illustrates recent attempts and designs of heterogeneous photocatalysts and the target molecule to which they have been tested.
As shown in Table 1, most of the reported systems involve the use of TiO2 as the photocatalyst. On the other hand, the use of supports has been varied encompassing inorganic materials (silica), organic supports, ashes and, derivate from wastes. From the heterogenization point of view, sol–gel seems to be the most widely employed method, probably due to its easiness of manipulation, versatility, and broad range of potential experimental variations. The use of approaches which provide nanostructurated systems has also been reported. Phenol, drugs, pesticides, and dyes have been investigated as the target pollutant.
|Catalyst system||Preparative method||Target molecule||Commentsa||References|
|MI-coated photocalaysts (P25 as precursor)||Organic polimerization||2,4-Dinitro|
|SF ranging from 1.59–3.29 for similar chlorophenols. CF ranging from 1.47 to 1.74 for similar chlorophenols. SF and CF were estimated from Kinetic experiments||(8)|
|MI TiO2/WO3 nanocomposites||Sol–gel||2-nitrophenol|
|Tetrabutyl Orthotitanate as precursor. SF ranging from 2.5 to 3.95. SF was estimated from adsorption experiments||(9)|
|MI based on fly-|
|Organic polimerization||Tetracycline||CF in ternary antibiotic solution ranging from 1.25 to 1.67. CF was estimated from Kinetic experiments||(10)|
|Acid-catalyzed sol–gel method||Carbamazepine, clofibric acid,|
|Pharmaceuticals and personal care products (PPCPs) pollutants. Photocatalytic activity of the novel TiO2–coconut shell powder (TCNSP) composite for the photodegradation of three PPCPs pollutants is investigated under UVC and black light blue UVA irradiation. Under the UVC/TCNP combination, 99% removal was achieved compared to 30% for P25||(11)|
|Impregnation Ti(OCH(CH3)2)4 solution was added to BEW solution||Atrazine||Synthesis of the TiO2 nanoparticles was performed by using BEW without a reducing agent and the photocatalytic activity of the catalyst was investigated for the degradation of atrazine with UV irradiation. TiO2 with waste material can be recycled and reused four times for the removal of atrazine. Degradation 85% after 70 min reaction||(12)|
|TiO2/almond shell activated carbon||Metal organic chemical vapor deposition and impregnation||Industrial phosphoric acid solution||Load TiO2 on the surface of active carbon by using two techniques. These photocatalysts were employed not only for the adsorption of total organic carbon (TOC) from industrial phosphoric acid solution. The catalysts showed performance higher than P25 with 90% degradation of the solution after 300 min||(13)|
|TiO2/perlite granules||Metal organic chemical vapor deposition and impregnation||Ammonia||Investigate the photocatalytical activity improvement of TiO2 (P25) powder as a photocatalyst which immobilized on perlite granules to remove ammonia from synthetic wastewater under UV irradiation||(14)|
|Sol–gel||Nitrogen monoxide (NO)||Modification of sand and fumed silica with titania in order to obtain a photocatalytic active material for the degradation of pollutants. The sample of sand showed a high photonic efficiency (0.41%) comparable to that of P25 (0.40%). The coated silica fumed containing the most titania content (28 wt%) showed the best photonic efficiency (0.56%) in the degradation of NO||(15)|
|TiO2/hydroxylated fly ash ceno|
|Sol–gel and photoreduction method||Methylene blue (MB)||TiO2 was coated on the surface of FACs by the sol–gel method. Platinum was then deposited on these TiO2/FAC particles by a photoreduction method. The photocatalytic activity for the degradation of MB under visible light irradiation. Photocatalytic experiments indicated that 3 g L−1 of the photocatalyst (calcinated 450°C with a Pt/TiO2 mass ratio of 1.5%) exhibited the best photocatalytic activity, as a 95% degradation of methylene blue after 8 h||(16)|
|TiO2/carbon composites||Self-assembly of carbon precursors and titanium isopropoxide||Rhodamine B||Pluronic F127 was employed as a soft template. The presence of titania nanoparticles uniformly dispersed in the carbon mesostructure. Synthesis of ordered mesoporous titania–carbon composites. Under UV radiation decomposition of RhB with TiO2/composites was (90–86%) and P25 was 50%||(17)|
|SnO2/carbon nanocomposites||Solvothermal method||Rhodamine B glyphosate||Photocatalytic activity on the degradation under simulated sunlight irradiation. Uniform distribution of SnO2 nanoparticles on the graphite-like carbon surface.||(18)|
|Au/TiO2 nanocomposite||Inverse miniemulsion by sol–gel||Rhodamine B||Visible-light photocatalytic activity in the degradation of the organic dye. The Au content in the nanocomposite particles could be conveniently tuned by the amount of HAuCl4 dissolved||(19)|
|TiO2/catalyst residue from a Ziegler–Natta catalyst petrochemical plant||Impregnation||Drugs||Drugs, namely, atorvastatin calcium, diclofenac sodium, fluoxetine, ketoconazole, ibuprofen, dexamethasone, tioconazole, naphazoline hydrochloride, valsartan, guaifenesin, and paracetamol. The highest drug degradation was observed under UV (48.6%) and visible (45.2%) radiation with the synthesized photocatalyst||(20)|
In parallel, some alternative approaches in the development of potential heterogeneous photocatalyst, namely (i) the use of industrial wastes as source of photocatalyst systems which were impregnated on commercial supports (20–22); (ii) the use of the concept of molecular imprinting (MI) to afford selectivity to the adsorption and photocatalyst degradation of target molecules (23–25), and (iii) combination of nano- and micro-particles as supports (26–28) for the impregnation of titania active species were explored. Regarding to the former, the rationale was that several photocatalyst systems imply the use of metal doping to guarantee a better photocatalyst activity or to shift the spectrum of actuation toward the visible spectrum region. The employed residues (industrial petrochemical chemistry, galvanic bath residuals, silver-containing residual solutions from photography classes, metal-containing residual solutions from academic analytical chemistry classes) contained several metals (Zn, Ag, Cr, Cu Mg, including Ti) which could exhibit photocatalyst activity (29), as shown in Scheme 1.
In the case of MI systems, template-shaped cavities are designed using polymer matrices with memory of the template molecules that can be applied in molecular recognition (30, 31). As reported in the literature (32, 33), matrices prepared by the sol–gel method offer certain advantages compared to conventional organic polymerization methods, such as rigidity, thermal stability, tailored porosity, and flexibility in processing conditions. MI process can be roughly described in Scheme 2. At first stage, a template–functional monomer complex is assembled via appropriate interactions. A cross-linking monomer is then used to form the solid MIP matrix. Finally, the removal of the template molecule liberates an imprint cavity of a defined size and shape.
The aim of such approach is to provide a selectivity factor for heterogeneous photocatalysts, considering that the target molecule would be employed in the synthesis of the solid material. The presence of the cavities in the shape of a given pollutant would favor its adsorption and photocatalytical decomposition viz-a-viz to the other species present in the milieu.
Concerning the third approach, nanosilicas are the materials of large surface area, bearing uniform and tunable pores, and small particle size and mechanical strength. These are important characteristics that may affect the photocatalytic activity and enhance the thermal stability of the photocatalyst. During the last decade, several mesoporous materials with different compositions of metal nanoparticles and complex nanostructured hybrid systems using SiO2 supports have been explored for degradation of RhB by photocatalysis. For instance, core@dual–shell SiO2–TiO2 composite fibers (34), TiO2/Au/SiO2 (~50 nm) (35), Fe3O4@SiO2@TiO2@Pt (~420 nm) (36), Pt@SiO2@TiO2 core–shell composites (~120 nm) (37), etc. But these catalyst systems do not exhibit very large surface areas, and they may present limitations in catalyst reuse due to their size. The proposed strategy here consists in combining silica nanoparticles (7.8–12.3 nm) produced by the sol–gel method with commercial microsilica as support for TiCl4 impregnation, which would be further calcinated to TiO2 (21, 24) as shown in Scheme 3. This system offers considerable potential as solid nano-/microstructured supports for the immobilization of heterogeneous photocatalysts overcoming the size problem in the reuse, generating catalyst species both on the surface of the nanoparticle moieties, as well as on the micrometric particle (silica bulk), which improve the adsorption and the contact between the pollutant molecules and the catalyst.
The aim of the present manuscript is to compare the photocatalyst activity of these three above-cited approaches, operating under the same reactor conditions and with the same target molecule (Rhodamine B). For comparative reasons, commercial TiO2 (Degussa P25) was also evaluated.
Silica solutions were prepared using tetraethoxysilane (TEOS, >98%, Sigma-Aldrich) and ethanol (99.5%, Nuclear) as the solvent. Hydrochloric acid (HCl, Nuclear, 38%) and ammonium hydroxide (29%, Nuclear) solution were used as catalysts by the sol–gel method. Silicon tetrachloride (SiCl4, 99%, Sigma Aldrich) was used to chemically modify the silica solutions. Sylopol-948 (Grace) was used as the micrometric silica reference. In addition, fumed silica (Wacker HDKN20), zeolite NaY (Zeolyst Internatial), and agroindustrial waste (rice husk) were used as the support for the photocatalysts. TiCl4 (Merck), metal-containing residual solutions from academic analytical chemistry classes and residual Ziegler–Natta catalyst slurry (Ti-based polymerization catalyst) from a petrochemical plant were employed for the preparation of the supported titania photocatalyst. Rhodamine B (C28H31N2O3Cl, PrótonQuímica, P.A) was used as the dye probe for the degradation tests. Double-distilled deionized water was used for the preparation of solutions used in the catalytic tests. For comparative purposes, Degussa P25 (denoted as TiO2) was also employed as received.
2.2. Preparation of supported photocatalysts
The molecular imprinting (MI) materials were prepared using the sol–gel process by acid route and TEOS as the raw material. Three different protocols were investigated: acid-catalyzed route 1 using TiCl4 as TiO2 precursor followed by a fast calcination (A1Ti), acid-catalyzed route 2 followed by a slow calcination (A2Ti), and acid-catalyzed route 3 using P25 as TiO2 precursor (A3P25). All samples were prepared with 0.2 M of hydrochloric acid at a 1:2 (HCl–TEOS) ratio. For each sample, RhB was fixed at 150 mg. RhB was added to a solution of TEOS followed by addition of TiCl4, and then the catalyst was added. For template removal, ultrasound-assisted (Branson-Sonifer®, Model 250) extraction was employed using methanol as the solvent (approximately 70 mL). These samples were labeled as molecularly imprinted (MI). Equivalent sample was prepared without the addition of RhB, that is, non-imprinted (NI) system (ATi). Both systems were submitted to a calcination process (450°C for 4 h at a rate of 5°C/min). In the case of system containing P25 (A3P25), this was added 2½ h after the beginning of the synthesis reaction. For comparison, the same material was synthesized without P25 (AWP25). There was no need for calcination in these systems. More details are reported elsewhere (23, 24, 38).
In the case of supported photocatalysts prepared using waste materials, 3 mL of residue was added to 1.0 g of fumed silica or zeolite NaY or rice husk. These materials were maintained under stirring for 90 min. Then, the solution was placed in a muffle furnace for 4 h for calcination (450°C). More details are reported elsewhere (20–22).
Functionalized nanosilicas were prepared by the hydrolysis of TEOS using a standard procedure described in the literature (21, 27). Thereafter, SiCl4 was added to the nanoparticles in solution. The silica nanoparticles were prepared using TEOS/SiCl4 in a molar ratio of 5.0–8.0. Hereafter, the resulting nanoparticles (powder) are labeled according to the TEOS/SiCl4 molar ratio employed in the synthesis. For example, SNCl5 refers to nanosilicas prepared with a TEOS/SiCl4 molar ratio equal to 5.0. Mixed-supported silicas were obtained by mixing the silica nanoparticles (previously synthesized) with 0.2 g of a Sylopol-948 (commercial micrometric silica (SM)) and stirred for 3 h. Mixed silicas were labeled as SMSNCl, indicating the presence of Sylopol-948 (SM) combined with synthesized nanosilicas (SNCl) from the preparation with SiCl4. Further details are reported elsewhere (27). 3.0 mL of TiCl4 was added to 1.0 g of synthesized silica particles and stirred for 90 min. After, the samples were then placed in a muffle furnace for 4 h for calcination (450°C). Two different nanometric supports were used in the production of nanometric catalysts (labeled SNCl5Ti to SNCl8Ti), and mixed silicas were used in the production of two mixed support catalysts (SMSNCl5Ti to SMSNCl8Ti).
Scheme 4 depicts the steps involved in the preparation of each supported photocatalyst as well as the corresponding labels. Preparation of MI materials is illustrated in Scheme 4a, b. Synthesis of functionalized nanosilicas catalysts and mixed-supported silicas ones is shown in Scheme 4e, d. Scheme 4e displays the preparation of catalysts using industrial and academics waste as the source of photocatalysts as well as the corresponding employed labels.
Table 2 presents the set of photocatalyst compared in the present study, as well as the corresponding label.
|TiCl4 + RhB||A1Ti|
|TiCl4 + RhB||A2Ti|
|P25 + RhB||A3P25|
|TiCl4 + SM||SMTi|
|TiCl4 + SNCl5||SNCl5Ti|
|TiCl4 + SNCl8||SNCl8Ti|
|TiCl4 + SMSNCl5||SMSNCl5Ti|
|TiCl4 + SMSNCl8||SMSNCl8Ti|
|Industrial and academic|
|Various metals + SiO2||SiVM|
|Petrochemical + SiO2||SiPe|
|Various metals + Zeolite NaY||ZVM|
|Petrochemical + Zeolite NaY||ZPe|
|Various metals + Rice husk||RHVM|
|Petrochemical + Rice husk||RHPe|
2.3. Characterization of the supported titania photocatalysts
The specific surface area (SBET), the pore diameter (DpBJH), and pore volume (VpBJH) were calculated by the Brunauer-Emmett-Teller method (BET) (39) and Barrett-Joyner-Halenda (BJH) algorithm (40), respectively. Measurements were taken using a Micromeretics TriStar II 3020 in the partial pressure range of 0.01 < P P0−1 < 0.95. Small-angle X-ray scattering (SAXS) measurements were performed at the SAXS1 beamline at the National Synchrotron Light Laboratory (LNLS), Campinas, Brazil. The scattered X-ray beam presenting a wavelength (λ) of 1.488 Å was detected on a Pilatus 300 k detector (27). I(q) vs. q scattering profiles were fit by multi-level of structural organization using the Beaucage model (41–43) in the Irena routine (44). The zeta potential of the nanoparticles and mixed supports (approximately 250 mg) was measured using Zetamaster equipment (Malvern Instrument). Metal content in the catalysts was determined by Rutherford backscattering spectrometry (RBS) using a He2+ beam of 2.0 MeV incident on the homogeneous tablets of the compressed (12 MPa) catalyst powder. SEM images were collected at the Center for Nanoscience and Nanotechnology (CNANO, UFRGS) in an EVO-50 (Carl Zeiss, Inc.) field emission scanning electron microscope equipped with a secondary electron detector Everhart-Thornley (ETSE) at 0.2–30 kV and with a working distance of 14 mm. UV–vis spectra of the supported photocatalysts in powder form were recorded using a Varian Cary 100 Scan Spectrophotometer with an accessory DRA-CA-301 (Labsphere) in the diffuse reflectance mode by co-adding 32 scans in the 200–800 nm range. The energy bandgap was determined by means of the Kubelka–Munk function. Scans ranged from 200 to 800 nm. Scanning electron microscopy (SEM) and energy dispersive X-ray (EDX) were performed using a JSM5800 (JEOL) microscope, operating between 5 and 20 kV. The samples were coated with a thin layer of conductive carbon by a sputtering technique. Diffuse reflectance infrared Fourier transmission spectroscopy (DRIFTS) was carried out at room temperature in a Bomem MB-102 spectrometer equipped with a diffuse reflectance accessory. Analyses at the absorbance modes were performed by co-adding 32 scans with 4 cm−1 of resolution. The spectral characteristic bands of the harmonic vibration of the silica materials network [νas(Si-O(-Si))] [νas(Si-O(-Si))] were studied by the FT-IR spectra in the region of 1300–1000 cm−1 (45–47). The broadband between 1300–1000 cm−1 was deconvoluted via Gaussian functions into four independent components: two (TO4 and TO6) associated with the transverse-optic (TO) modes and two (LO4 and LO6) with the longitudinal-optic (LO) modes of Si-O-Si using a nonlinear least-squares fitting method . The percentage of sixfold rings (SiO6) in the silica network (Q6) was estimated using the following ratio of fitted areas (27, 48, 49):
2.4. Statistical analysis
The SPSS Statistical software (SPSS for Windows, version 19, IBM®) was used to analyze the relationships among the data. All statistical tests were performed at the P < 0.05 level of significance. The Spearman correlation coefficient () was used as a nonparametric measure of statistical dependence between two variables.
2.5. Photocatalyst tests
For tests with the Rhodamine B, 0.0175 g of catalyst was added to 25 mL of the test molecule (20 mg L−1) at pH of the reaction medium (pH ≅ 4.3). The solution was then transferred to the reactor, and after adjusting the temperature (at 30°C), the lamp was switched on to start the irradiation. During each experiment, circulation of suspension was maintained to keep it homogenous and to have uniform temperature. Samples (5 mL) were taken at regular time intervals (0, 5, 15, 30, and 60 min) and then centrifuged (Cientec CT-5000R) for 20 min at 5000 rpm in order to separate the catalyst particles from the samples. To determine the concentration of RhB, solution absorbance was read in a Varian Cary 100 UV–vis spectrophotometer at the wavelength of maximum absorbance of target molecule (553 nm for RhB), and the absorbance was related to the concentration of the RhB through a calibration curve: Abs = 0.2053 C (mg L−1) (R2 = 0.9911; N = 7). All photocatalytic tests were performed in duplicate (error value lower than 5%). Photolysis tests were also performed to determine the percentage of dye degradation due to UV light exposure without the presence of a photocatalyst.
3. Results and discussion
According to the RBS results, the metal concentration is low in terms of the Ti/Si ratio. Systems SMSNClTi showed a Ti average composition of 0.45 wt%/SiO2, whereas the SNClTi obtained 1.33 wt%/SiO2. Therefore, the SN5ClTi achieved a higher metal content. According to the SEM-EDX results, the average atomic percentage of Ti was 16.04 ± 3.05 at.%. As shown in Table 3, Si was present in all the samples. Depending on the photocatalysts from wastes, Cu, Zn, Al, Ti, Cr, or Mg were also present.
Photocatalyst activity in the degradation of rhodamine B was comparatively evaluated by the three supported catalyst systems, under UV and visible irradiation, as shown in Table 4.
As shown in Table 4, the efficiency of the photolysis degradation was <8%. The P25 showed an excellent performance in the UV (82%), but poor performance in the visible (12.3%). All of the supported titania photocatalysts exhibited better photocatalyst activity (25.4–61.6%) for RhB degradation than commercial P25 under visible radiation. The SNCl5Ti catalyst showed the best photocatalyst activity under UV (68%) irradiation, followed by SiPe (66.8%).
The approach of nanometric photocatalysts showed better performance (59–68%) than the nano-micrometer systems (33.6–42.7%) under UV and visible radiation. This is due to the incorporation of higher amount of titanium on the surface of nano-catalysts, as shown in the results of RBS. However, both approaches are valid for the degradation of organic pollutants from supports such as silicas synthesized by sol–gel, because nano-micrometer photocatalysts (25.4–30.1%) also were better than the P25 under visible radiation. The advantage of a nano-micrometric system resides in the possibility of reusing the catalyst several times as have already studied (27, 28, 50), which is a current limitation of nanoscale catalysts.
All the supported photocatalysts from industrial waste exhibited photocatalyst activity in dye degradation, although with catalyst activity lower than that of Degussa P25 under UV radiation. On the other hand, all the catalysts from industrial and academic wastes exhibited activity higher than that of Degussa P25 under visible radiation, suggesting a potential application of such photocatalysts under sun irradiation. Among them, the SiPe catalyst showed the best photocatalyst activity under UV (66.8) and visible (61.6%) irradiation. The higher efficiency in the photocatalytic activity of the SiPe sample compared with the other samples may be because the SiPe sample has the lowest Eg value, the largest ZP value (in the module), and the highest SBET (available surface area of the active sites) and Vp values.
|Method||System||Degradation RhB (%) radiation|
|Molecular imprinted (MI)||A1Ti||32.2||n.d.|
|Nanoparticles and mixed silicas||SMTi||n.d||n.d.|
|Photocatalysts from wastes||SiVM||41.1||48.2|
We have shown that our MI systems showed selectivity and competitiveness up to 1.9 and 3.5, respectively, toward RhB. Also, compared to the commercial sample (P25), selectivity (up to 180%) and competitiveness (up to 290%) were obtained (24). However, the average of activities under UV radiation was lower than the average values obtained by nano-silica (1.44-fold lower) and industrial wastes routes (1.16-fold lower). On the other hand, the lowest value of activity achieved within the MI systems containing TiO2 (32.2%) was comparable to the SMSNC18Ti (33.6%) and higher than RHVM (27.8%), which showed the lowest activities within nanosilica and industrial wastes routes, respectively. Despite their lower activities, one of the main advantages of MI approach is the possibility to achieve better selectivity removal compared to the commercial samples and also reusability (51), which are recognized drawbacks in heterogeneous photocatalysis (6). Figure 1 shows the results of RhB photocatalyst decomposition under visible and UV radiation by selected photocatalysts from each of the three approaches.
As shown in Figure 1, comparing among all the supported photocatalysts, the performance of P25 is inverted under visible radiation, becoming the lowest among all. The catalysts from industrial waste dominate over nanosilicasand mixed silicas supports.
In order to understand the differences concerning each alternative route, these systems were characterized by a series of complementary techniques aiming at accessing some information concerning textural, structural, and morphological aspects of these supported catalysts.
3.1. Textural characteristics of photocatalysts
The catalytic activity of the particle samples is proportional to their specific surface area because it requires the adsorption of reactants and also the transfer of photoexcited electrons into the adsorbed molecules, which is directly related to the diffusion of the RhB to the catalyst center (52). The specific surface area (SBET), the specific pore volume (VpBJH), pore diameter (DpBJH) are presented in Table 5.
|Molecular imprinted (MI)||A1Ti||634.5||0.47||3.1||9.8||3.4||−22.9||2.1||48.7|
According to Table 5, all systems prepared are mesoporous materials as classified by the International Union of Pure and Applied Chemistry (IUPAC) (53, 54). It can be seen that all systems have shown values of surface area higher than the commercial sample. Also, the mean value achieved for the MI systems is ca. 4.6-fold and 3.1-fold higher than the mean values found for nanometric and photochemical residue, respectively. Moreover, the supported photocatalysts with TiCl4 showed a reduction in the final surface area in relation to the nanometric and mixed supports, which also are observed in photocatalysts prepared with petrochemical residue and solution of various metals supported on zeolite NaY (ZVM and ZPe) and the system with molecular imprinted in which P25 was added (A3P25). This fact can be due to the high calcination temperature providing the agglomeration of the catalytic metal particles, thus decreasing the available surface area (55, 56) and the degree of structural organization and internal pores between nanoparticles, mixed silicas, and zeolite NaY (27, 56). Considering the resulting heterogeneous photocatalysts from TiCl4 or wastes impregnation, the SiPe reached the highest specific surface area and pore volume. This increase can be associated with action of the waste or TiCl4 on the respective support rendering difficult the contact between the particles and avoiding the agglomeration of the catalytic particles.
3.2. Multi-scale organization of the photocatalysts analyzed by SAXS
The photocatalysts were further analyzed by SAXS, which provides structural information, that is, a broader and clearer view of catalysts formation and aggregation (57). The multi-hierarchical organization of these materials, their gyration radius (Rg), and fractal structure (P) were investigated through the Unified approach (41–43) in the Guinier and Power-law regions. Table 5 shown radius of gyration of particles (Rg) and the linear Power-law decay (P), obtained through Unified fit on SAXS profiles of photocatalysts.
As shown in Table 5, photocatalysts are organized into three levels, excepting ATI and P25 which show two levels. Rg results show that the primary particle size ranges from 0.6 to 12.3 nm, according to the preparation method. Catalysts SMSNCl5Ti was the largest particle, due to the combination of nanoparticles incorporated into the surface of a micrometric support as (SM), resulting in bigger aggregated mixed silica catalyst (SMSNClTi).
Comparing fractal structure (P) results, P25 is a completely spherical particle, as well as the photocatalyst from industrial wastes (P = 4), which corresponds to spherical and dense clusters of approximately smooth surface according to Beaucage (42, 43). On the other hand, all MI systems showed fractal surface (3.3 < P < 3.7), whereas that nano- and micronano-photocatalysts mass fractals (2.0 < P < 2.5) to surface fractals aggregates (P = 3.6).
3.3. Zeta potential (ZP) measurements
Zeta potential (ZP) measurements were used to verify the stability of the suspensions of the studied systems, to determine its influence on the RhB dye adsorption of the supported photocatalysts and to correlate the electric potential information with the results obtained by SAXS. Table 5 shown the ZP and Eg values. Figure 2 shows the relationship among SBET, Rg and ZP of the best catalysts for each strategy discussed in this article.
The supported catalysts using the MI approach present ZP between −10.6 and −22.9 mV and in the case of synthesized nanoparticles and mixed silicas, ZP lies between −10.8 and −29.5 mV. For the impregnated with industrial and academic wastes ZP was between −20.3 and −36.4 mV (as shown in Table 5).
According to Figure 2a, there is a direct correlation (> 0.705) between electronic (ZP) and textural properties (SBET) of photocatalysts. Among all the photocatalysts, SiPe showed the highest ZP (−36.4 mV) (in module) and all systems bear a negative charge on their surfaces. The negative charge density on the surface is greater in the smaller photocatalyst, most likely due to the smaller particle size (2.4–6.8 nm). It is worth noting that high surface area supports are desirable for photocatalysis because it can promote the adsorption of the dye RhB to be degraded on the catalyst/support surface (21). Such behavior is in accordance with the results of Ribeiro et al. (58) and Gaya and Abdullah (59), who reported that the performance of a photocatalyst is dependent on the surface textural characteristics (size, surface area, and pore volume) because the local morphology strongly influences the number of striking photons and the rate of the photocatalytic reaction.
Conversely, it was possible to observe a strong direct correlation (> 0.757) between the zeta potential and the particle radius supported catalysts, as shown in Figure 2b. A smaller particle size corresponds to a higher negative surface charge: The nanometric photocatalyst (SiPe, ZP = −36.4 mV and Rg = 2.7 nm) exhibited this trend. The preparation of photocatalyst by impregnation with industrial and academic waste showed two effects: (i) increase the particle radius (Rp) and (ii) enhance the negative zeta potential. In this study, better photocatalyst showed an increase in zeta potential as the particle radius increased (−10.9 to −36.4 mV). The trend is consistent with the fact that the van der Waals forces are dominant over the electrostatic repulsive ones: the electrical double-layer thickness decreases, thereby promoting the agglomeration of primary particles, which in turn affords the particle growth size, as shown in the SAXS analyses.
According to Figure 3, the performance of heterogeneous photocatalysts under UV radiation is related to the zeta potential showing a strong correlation (> −0.827). In the present case, RhB is characterized by its cationic nature, and its compatibility with the photocatalyst surface potential (negative charge) may potentially improve the photocatalyst activity (59). On the other hand, all the photocatalysts exhibited activity higher than that of Degussa P25 under visible radiation, suggesting a potential application of such photocatalysts under sun irradiation. It is important to highlight that SNCl5Ti is synthesized from silicas nanoparticles, whereas SiPe is prepared from petrochemical waste. Both strategies are viable for the degradation of organic pollutants. Based on the ZP results, we established that photocatalyst dispersions are stable and that the BET and SAXS analyses are in agreement and interrelated with the discussion about the electronic properties of these photocatalysts.
3.4. Energy bandgap (Eg) by DRS
The energy bandgap (Eg) is a relevant parameter for the photocatalytic process because lower values of Eg correspond to less radiation energy required to activate the process, which may reduce the required radiation into the visible light range (60). To determine the photoabsorbance properties, UV-DRS was employed in the wavelength range of 200–800 nm. Table 5 shows Eg of photocatalysts. For comparative reasons data from commercial photocatalyst (P25) were also included.
It was possible to observe strong correlations (> −0.749 and > −0.788) between the energy bandgap (Eg) with pore volume (VpBJH) and surface area (SBET). Thus, the higher the pore volume, the higher the surface charge (in module), which may promote diffusion of the dye from the solution to the catalyst centers. On the other hand, the higher the surface charge (in module), the lower the energy to promote the electron from the valence band to the conduction band. Comparing the Eg values between the prepared photocatalysts, as shown in Table 5, the Eg of the catalysts ranged from 1.8 to 3.7 eV. Notably, the SiPe sample exhibited the lowest Eg value among the prepared titania photocatalysts. These results illustrate the role played by Eg on the photocatalytic process and the effects of textural (surface area and pore volume) and electronic (ZP) characteristics.
3.5. Four- and six-membered Silica
Comparing MI systems with nanoparticles and mixed silicas, one can see that systems containing TiCl4 (A1Ti, A2Ti and ATi) and the systems with lower TEOS:SiCl4 ratio (SNCl5Ti and SMSNCl5Ti) afford the lowest %(SiO)6 content (below than 55%). Thus, these materials tend to present more rigid structure of silica represented by four-membered silica ((SiO)4). In the other words, both the presence of P25 and higher TEOS:SiCl4 ratio result in more hydrophilic materials. As a general trend, hydrophilic materials seem to be more easily achieved by a protocol using the synthesis of silica nanoparticles than the MI materials. Although %(SiO)6 was not correlated with degradation of RhB, it is interesting to note that materials with the highest %(SiO)6 content (A3P25 and SNCl18Ti) afforded better degradation under UV radiation.
3.6. Morphological characteristics obtained using SEM
Scanning electron microscopy (SEM) was used to evaluate the particle morphologies of supports. The low- and high-magnification images of the nanoparticles are shown in Figure 4.
According to Figure 4a, b supports SPe and ZPe are larger particles and approximately rounded, whereas synthesized silica supports (SMSNCl5Ti) are smaller, amorphous, and aggregate particles (Figure 4d). In the case of MI materials, as observed in Figure 4c, a large number porous and some particle agglomerates are present on the surface of system. In the case of MI materials, as observed in Figure 4c, a large number porous and some particle agglomerates are present on the surface of system.
This work presented three alternative approaches for the degradation of organic pollutants such as RhB. Imprinting molecular materials, synthesized silica particles as catalytic supports from TiCl4 and photocatalysts from industrial and academic waste demonstrated activity for degradation of RhB dye under visible (61.6%) and UV (68%) radiation and are better than the commercial TiO2 (12.3%) under visible radiation. The textural and structural characteristics of the supported catalyst prepared with fumed silica and petrochemical waste (SiPe), namely the low-energy bandgap (1.8 eV), large surface area (280 m2 g−1), high pore volume (1.9 cm3 g−1), and high zeta potential value (−36.4 mV), may have been responsible for their high activity.
Our alternative approaches herein studied could help to overcome the main drawbacks involved in heterogeneous photocatalyst, such as the lack of selectivity, the low surface area, and UV-limited spectrum for commercial samples.
This project was partially financed by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). Yolice P. Moreno, Cícero C. Escobar and William L. da Silva are grateful for the grant provided by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES). The Brazilian Synchrotron Light Laboratory (LNLS, Campinas, Brazil) and the Laboratory of Ion Implantation are thanked for the analysis of the SAXS (Projects D11A-SAXS1-8691 and SAXS1-15911) and RBS, respectively.