Synthesis, characterization, and photocatalytic activity of multicomponent CdMoO4/g-C3N4/GO composite
Article information
Abstract
CdMoO4/g-C3N4/GO composites were first synthesized by a hydrothermal process for the photocatalytic degradation of rhodamine B (RhB) under visible light illumination. The findings show that CdMoO4 is highly dispersed onto g-C3N4 and graphene oxide (GO) sheets. The surface area of the composite increased 27–30.5 times that of CdMoO4, and its band gap energy decreased by about 1.26 times. These features significantly improve the photocatalytic activity of the composite in the RhB decomposition reaction under visible light. The RhB degradation efficiency of the CdMoO4/g-C3N4/GO composite is 5.7, 1.36, and 1.65 times that of CdMoO4, CdMoO4/g-C3N4, and CdMoO4/GO composites, respectively. The active species trapping experiments show that the main forms in RhB degradation are •O2−, •OH, and h+. The stability of the photocatalyst is retained even after the 5th reuse. In addition, RhB degradation products were identified with the high-performance liquid chromatography-mass spectrometry method, and the pathway of photocatalytic degradation was also addressed.
Abstract
Graphical Abstract
1. Introduction
Currently, along with the increase in industrial production activities, the emission of hazardous organic substances into the environment, especially water bodies, is increasing [1,2]. Hazardous organic pollutants, such as dyes released into water bodies, contaminate water sources, affecting the health of humans, animals, and plants that depend on the water source. These dyes are reported to be toxic, carcinogenic, and mutagenic [3,4]. Therefore, removing toxic organic compounds from polluted waters to protect public health is an increasingly important and urgent issue. Various methods such as adsorption [3,5], biological degradation [6–8], and advanced oxidation processes (AOPs) [9–13] are developed to removal of pollutants from wastewater. Among AOPs, photocatalysis acted in the visible-light region has been acknowledged as a promising alternative for removing toxic organic pollutants in wastewater because it has various outstanding advantages, such as using abundant energy sources from sunlight, high reaction rate, low energy consumption, mild reaction conditions, high stability, and environmental friendliness [14,15]. For these reasons, numerous photocatalysts have been synthesized and evaluated for their ability to degrade organic pollutants. Photocatalysts can be mentioned as common pure semiconductors, such as TiO2 [16–18], ZnO [19–21], and SnO2 [22–24] or multicomponent photocatalysts, such as ZnO/CuO [25,26], TiO2/C [27], TiO2/Cu2O [28], CuO/TiO2/ZnO [29], Ag3PO4/TiO2 [30], and AgBr/Ag3PO4 [31]. Among these semiconductors, single-component systems show poor photocatalytic performance in the visible-light region because they have wide-bandgap energies and fast recombination between electrons and photogenerated holes. In contrast, multicomponent photocatalysts exhibit a better performance towards organic pollutant decomposition under visible light thanks to the formation of heterojunctions between the semiconductors. Consequently, they reduce the bandgap energy and hinder the photogenerated electron-hole recombination. Therefore, the synthesis of multicomponent semiconductor composites to improve the photocatalytic activity in hazardous organic substances decomposition under visible light is a potential research direction.
Over the past few years, several studies have suggested that CdMoO4, a semiconductor with a scheelite-type structure, is a potential photocatalyst [32–34] thanks to its unique combination of chemical and physical properties such as pressure-induced phase transformations [35], electronic excitation at vacuum ultra violet synchrotron radiation [36], electrical structure, optical properties [37] and relatively low band gap energy as compared to other scheelite materials [34,38]. However, with wide-bandgap energy (~3.42 eV), pure CdMoO4 only exhibits good photocatalytic activity under ultraviolet radiation [32,33,39]. To enhance the photocatalytic activity of CdMoO4 in the visible light region, researchers have synthesized its multicomponent composites such as AgI/CdMoO4 [40], CdS/CdMoO4 [41], and Bi-Doped CdMoO4 [42].
Graphitic carbon nitride ( g-C3N4) is a layered polymer semiconductor with high thermal stability, easy synthesis, low cost, and nontoxic nature [43,44]. With its narrow bandgap energy (2.7 eV), g-C3N4 can absorb light with wavelengths up to 450 nm, thus showing potential as a photocatalyst in the visible region [45,46]. Therefore, the combination of CdMoO4 and g-C3N4 to form heterojunctions creates interlacing bandgap energy edges, increases the absorption of visible light, and prevents the recombination of electrons and holes, thereby increasing the photocatalytic activity of the composites [39,43,44]. However, g-C3N4 has a small surface area and almost no electrical conductivity [45,46]. The shortcomings of g-C3N4 limit the use of the composites. These restrictions can be overcome by combining g-C3N4 with GO or reduced graphene oxide (rGO). GO and rGO possess a layered structure similar to g-C3N4, made from 2D carbon sheets. Furthermore, GO and rGO have a large surface area, high mechanical and thermal stability, and excellent optical properties [47,48]. Thus, these materials are a perfect complement to overcome the limitations of g-C3N4 in the photocatalysis field. Recent studies have shown that with the addition of GO or rGO, the g-C3N4-based material exhibits enhanced photocatalytic activity. For example, Pawar et al. [49] reported that the ability to degrade RhB of the g-C3N4/CdS/rGO composite under visible light was 3.4 and 2.35 times that of g-C3N4/CdS and rGO/CdS, while the degradability of congo red of g-C3N4/CdS/rGO was 6.0 and 4.38 times that of g-C3N4/CdS and rGO/CdS. Wang et al. [50] also claimed that the ternary component composite GO/Ag3PO4/g-C3N4 could catalyze the degradation of RhB more significantly than the binary component composite g-C3N4/Ag3PO4 or GO/Ag3PO4, as well as each component alone under visible light. Another study by Huang et al. [51] revealed that the g-C3N4/TiO2/rGO composite exhibited a much better ability to degrade methyl orange, RhB, and phenol than g-C3N4/TiO2 or TiO2/rGO. Therefore, it can be seen that the simultaneous dispersion of CdMoO4 on the g-C3N4 and GO or rGO substrates can further improve the photocatalytic activity of the obtained material under visible light.
To the best of our knowledge, few articles on trinary component photocatalysts have been reported. Therefore, in this study, we synthesized a CdMoO4/g-C3N4/GO composite photocatalyst with the hydrothermal method at low temperatures. The combination of CdMoO4, g-C3N4, and GO is expected to obtain a composite with unique physicochemical properties in terms of surface area and band gap energy. The photocatalytic activity of the synthesized materials was investigated via the decomposition of RhB solution under visible light. It is well-known that RhB is a dye of the xanthene family, used in different fields, especially textile dyeing, and it is frequently detected in effluents. RhB is toxic to humans and animals, causing skin, eye, and respiratory irritation. It can also be teratogenic, carcinogenic, and mutagenic [52,53]. Therefore, RhB was selected as a dye model for degradation in this study. The influence of CdMoO4 content on the composite’s photocatalytic performance was investigated. The RhB degradation ability of CdMoO4/g-C3N4/GO composites, single materials (CdMoO4, g-C3N4, GO), and binary component composites (CdMoO4/GO or CdMoO4/g-C3N4) was compared. In addition, the kinetics and photocatalytic mechanism of RhB degradation were also proposed.
2. Experimental
2.1. Materials
Graphite powder, KMnO4, NaNO3, H2SO4, Cd(NO3)2×4H2O, Na2MoO4×2H2O, and RhB were purchased from Merck. Urea, H2O2, and HCl were purchased from Guangzhou Company, China. All chemicals are of analytical grade.
2.2. Preparation of Graphene Oxide, Carbon Nitride Sheet, and CdMoO4/g -C3N4/GO Composite
Graphene oxide (GO) was synthesized from graphite powder, according to Hummers’ method [48]. Mix 1 g of graphite and 0.5 g of NaNO3 in a 1-litre beaker. Add 23 mL of 98% H2SO4 solution to the above mixture; stir for 5 min at 5°C. Slowly add 3 g of KMnO4 to the beaker; continue stirring for 2 h, then increase the temperature to 35°C and stir for another 30 min Add 46 mL of H2O to the system and gradually increase the temperature to 98°C, stirring for 30 min until the black mixture turns brown. Add 140 mL of H2O and 10 mL of 10% H2O2 further to the mixture. Centrifuge to separate the product and wash with a 5% HCl solution. Then, wash with H2O at 70°C. Finally, centrifuge, separate the solids, and dry at 60°C to obtain GO.
g-C3N4 was synthesized by pyrolyzing urea [54]. Urea was dried at 80°C for 2 h, then raised to 550°C and kept for 3 h. The product was washed with a 0.1 M HNO3 solution and distilled water to pH 7 and dried at 80°C for 24 h.
The CdMoO4/g-C3N4/GO composite was synthesized with the hydrothermal method [39,50]. Disperse 0.04 g of GO in 50 mL of distilled water and sonicate the mixture for 45 min. Next, add a calculated amount of Cd(NO3)2×4H2O to the solution and stir for 3 h (300 rpm) to obtain suspension 1. Add 0.08 g of g-C3N4 to 50 mL of distilled water and sonicate for 45 min. Then, add a quantity of Na2MoO4×2H2O (the number of moles corresponding to the number of moles of Cd(NO3)2×H2O) to the solution and stir for 3 h (300 rpm) to obtain suspension 2. Add suspension 2 dropwise to suspension 1 and adjust the system to pH 5 with a 0.5 M HCl solution. Place the mixture in a Teflon flask and conduct the reaction at 120°C for 1.5 h. Finally, centrifuge, wash the precipitate to pH 7, and dry the solid at 80°C to obtain CdMoO4/g-C3N4/GO. The composites synthesized with 10, 20, 30, and 40 weight per cent of CdMoO4 are denoted M1, M2, M3, and M4, respectively. The schematic diagram of the process is illustrated in Fig. S1.
2.3. Characterization
Powder X-ray diffraction (XRD) used for determining the phase structure of synthesized materials. XRD patterns was recorded on the VNU-D8 Advance diffraction machine (Bruker, Germany) with a CuKa radiation source at a wavelength of 1.5406 Å and scanning angle of 5–80°. Thermogravimetric (TG) was used to determine the CdMoO4 content in the composites. It was performed on a Labsys TG/DSC1600 (Setaram) instrument from ambient temperature to 900°C in air. The morphological features were analyzed from the images taken with a field emission scanning electron microscope (FE-SEM, S-4800, Hitachi) operated at 5 kV. Fouriertransform infrared spectra (FTIR) within 400–4000 cm−1 were recorded on an IR-Prestige-21 (Shimadzu) spectrophotometer using the pelleting technique with KBr. FTIR is used to characterize the functional groups as well as bonds in synthesized materials. Energy-dispersive X-ray (EDX) spectra and EDX-mapping were recorded on a Cary 5000 (Agilent) and used to determine the surface composition and distribution of elements in the composites. Nitrogen adsorption-desorption isotherms were measured on a Micromeritics Tri Start 3000 instrument. Samples were degassed at 250°C for 5 h before measurement. Specific surface area was calculated using the Brunauer–Emmett–Teller (BET) model. Ultraviolet-visible Diffuse Reflectance Spectroscopy (UV-Vis-DRS) were recorded on an Aligent Cary 5000 spectrophotometer with BaSO4 as reference. UV-Vis-DRS was used to evaluate the optical properties of the synthesized materials. Metal analysis dissolved in the solution was performed by the Atomic Absorption Spectroscopy (AAS) using a PerkinElmer Analyst 200. The identification of reaction products was performed by using HPLC-MS (high performance liquid chromatography–mass spectrometry (LC-MS Agilent 1100) under the following conditions: Zobac C18 column with a particle size of 5 μm, water + 0.1% formic acid and acetonitrile as a mobile phase, a flow rate of 0.4 mL/min, an injection volume of 10 μL, the atmospheric pressure electrospray ionization (AP-ESI) source, and operated in positive mode. Chemical Oxygen Demand (COD) was determined by the Dichromate method, using ECO 25 thermoreactor (Velp) and Shimadzu-UV-Vis spectrophotometer (model 1240).
2.4. Photocatalytic Activity Measurements
RhB was used to evaluate the photocatalytic activity of the synthesized composites. An amount of RhB dye (10 mg) was dissolved in distilled water (1 L) to prepare the initial solution (10 mg/L). The measurements were carried out as follows: 25 mg of the photocatalyst was dispersed in 125 mL of 10 mg/L RhB aqueous solution. Then, the suspension was illuminated with a xenon lamp (250 W) equipped with a UV cutoff filter. Before irradiation, the suspension was stirred in the dark for 60 min at 30°C to achieve adsorption-desorption equilibrium. At different time intervals, a volume of the sample was withdrawn, centrifuged to remove the photocatalyst, and the concentration of RhB was determined on a Shimadzu-UV-Vis spectrophotometer (model 1240) at a maximum wavelength of 554 nm (λmax). The RhB degradation efficiency was calculated according to Eq. (1).
where C0 and Ct are the RhB concentrations (mg/L) before and at illumination time t.
Since the λmax of some dyes is dependable in pH. Both the pH and the λmax of RhB during reaction were recorded (Fig. S2). It was found that the pH tends to decrease slightly from 5.93 for initial solution (10 mg/L) to 5.03 for the solution after 120 min of reaction and the λmax of RhB seems to unchangeable at around 554 nm. Hence, the calibration curve with λmax of 554 nm is used to measure the RhB concentration.
In order to evaluate the role of free radicals in the photocatalytic process of RhB degradation, KI (h+ trapping), isopropanol (IPA) (•OH trapping), and benzoquinone (BQ) (•O2− trapping) were used as radical scavengers [55]. The radical scavenger concentration is 5 mM and was added to the RhB solution before irradiation.
To develop a calibration curve for RhB concentration versus the absorbance, we prepared RhB solutions with different concentrations (1–15 mg/L). The absorbance of the solutions at the maximum wavelength of 554 nm was recorded on a Shimadzu-UV-Vis spectrophotometer (model 1240).
3. Results and Discussions
3.1. Characterization of Prepared Composites
Fig. 1a presents the XRD patterns of the g-C3N4, GO, and CdMoO4/g-C3N4/GO composite prepared with different CdMoO4 contents. g-C3N4 exhibits two characteristic peaks at 13.1 and 27.4°, corresponding to the graphitic layered carbon nitride structure. GO displays its typical peak at 11.3° [50]. All the composite samples show the peaks characteristic for the CdMoO4 phase at 29.2, 32.0, 34.8, 47.9, 50.0, 55.2, 58.9, 60.5, 73.4, and 77.8°, corresponding to diffraction planes (112), (004), (200), (204), (220), (116), (312), (224), (400), and (316) (JCPDS 07-0209) [33,39]. Besides, the composites also exhibit two broad peaks at about 11.3–13.1° and 27.4°, typical for the layered structure of g-C3N4 and GO. When the CdMoO4 content increases, the intensity of the peaks representing CdMoO4 also increases, and that of the peaks representing the matrix decreases. This result is consistent with what Chai et al. [43] reported when dispersing CdMoO4 onto g-C3N4. Thus, it can be assumed that CdMoO4 disperses into the g-C3N4 and GO matrix.
XRD patterns (a) and N2 adsorption-desorption isotherms (b) of GO, g-C3N4 and CdMoO4/g-C3N4/GO samples with different CdMoO4 contents.
The nitrogen adsorption-desorption isotherms were used to determine the porous properties of the synthesized materials. As can be seen in Fig. 1b, all studied samples have an isotherm of type IV with hysteresis loops of type H3 according to the IUPAC classification. This indicates that these materials have a porous structure, with the pore distribution curve shown in Fig. S3. As can be seen in the figures, the composites contain multi-capillaries with a pore distribution in the range of 2–100 nm, peaking at 4 and 38 nm. However, their specific surface area and pore volume are different (Table S1). The specific surface area of the GO, g-C3N4, CdMoO4, M2, M3, and M4 samples is 161.6, 75.2, 1.4, 42.9, 41.4, and 37.8 m2/g, corresponding to a pore volume of 0.73, 0.59, 0.01, 0.24, 0.29 and 0.2 cm3/g, respectively. The obtained composites have a smaller specific surface area than the matrices (GO, g-C3N4), but much larger than CdMoO4. Two reasons for the surface area reduction of CdMoO4/g-C3N4/GO can be considered. First, some micropores of g-C3N4 and GO were occupied by CdMoO4 particles during the synthesis. Second, some parts of the CdMoO4/g-C3N4/GO composite powders aggregated after CdMoO4 particles were incorporated into the composite [56,57]. The increasing content of CdMoO4 slightly reduces the surface area of the composites. However, these CdMoO4/g-C3N4/GO composites have a much larger specific surface area than the CdMoO4/g-C3N4 composites previously published [39,43], expecting an improved photocatalytic performance.
Fig. 2 presents the SEM images of GO, g-C3N4, CdMoO4, and composite (M3) samples, EDX spectrum and elemental maps of the M3 sample. It can be seen that GO has the form of thin and wide sheets stacked on each other (Fig. 2a). g-C3N4 also has stacked sheets that are thicker and narrower than the GO sheets (Fig 2b). These two materials exhibit a clear porous structure. The SEM images of the CdMoO4 sample with different magnifications in Fig. 2c show that this sample contains pea-like nanoparticles with a size of 50–150 nm, arranged like a pistil with a diameter of 1–2 mm. It can be observed in Fig. 2d that the composite consists of interwoven sheets of GO, g-C3N4, and CdMoO4 nanoparticles. The EDX spectrum (Fig. 2e) shows that all elements, namely Cd, Mo, O, C and N are present in the composite. The element maps also show a uniform distribution of the elements (Fig. 2f–k). The FTIR spectra of CdMoO4, g-C3N4, GO, and CdMoO4/g-C3N4/GO in the 4000–400 cm−1 region are presented in Fig. 3a. For GO, the absorption bands at 3414 cm−1 and 1630 cm−1 are attributed to the stretching and bending vibration of the O-H bond of the hydroxyl group or adsorbed water [58]. The bands at 1719 is assigned to the stretching vibration of C=O in the carboxyl group [58,59]. Other broad absorption bands at 1225 and 1055 cm−1 belong to the stretching vibration of epoxy and alkoxy C-O [59,60]. The peak at 1383 cm−1 presents the C-O stretching vibrations of carboxyl groups [58]. The FTIR spectrum of pure CdMoO4 exhibits a strong band at 777 cm−1 and a weak peak at 438 cm−1, representing the stretching and bending vibration of the Mo-O-Mo bonds in the MoO42− tetrahedron group. In addition, a broad peak at 3460 cm−1 and two sharp peaks at 1630 and 1395 cm−1 on this spectrum correspond to the valence and bending vibrations of the O-H bond of adsorbed water [43,44]. For the g-C3N4 sample, the FTIR spectrum shows the presence of a sharp characteristic peak at 810 cm−1, assigned to the stretching vibration modes in the triazine units. The bands at 1242, 1323, 1416, and 1570 cm−1 are related to the stretching vibration of heterocyclic C–N bonds. The absorption peak at 1630 cm−1 arises from the C=N stretching vibration, while the broad absorption band at 3258 cm−1 can be attributed to the N-H stretching vibration of the amino groups [44,58,61]. In the CdMoO4/g-C3N4/GO FTIR spectrum, all characteristic peaks of each component are observed. These peaks, however, are slightly shifted. The peak belonging to the stretching vibrations of the Mo-O-Mo bond occurs at 770 cm−1, while the peak corresponding to the stretching vibration of the triazine units is observed at 812 cm−1. Those representing the oxygen functional groups in GO and the CN functional groups in g-C3N4 in the range 1200–1700 cm−1 or peaks representing the O-H and N-H bonds in the range 3000–3500 cm−1 also have a slight shift in the wavenumbers. This shift may be due to the interactions between the phase components in the resulting composite.
SEM images of GO (a), g-C3N4 (b), CdMoO4 (c) and M3 (d); EDX spectrum of the M3 sample (e); Elemental maps of the M3 sample: Cd (f), Mo (g), O (h), C (i), and N (k).
FTIR spectra (a), TG diagram (b), UV-Vis-DRS spectra (c) band gap energies (d) of CdMoO4, g-C3N4, GO and CdMoO4/g-C3N4/GO samples.
To determine the CdMoO4 content in the composites, we analysed their TG recordings. The TG plot in Fig. 3b shows that the g-C3N4 and GO samples completely burn when the temperature rises to 700°C, while the CdMoO4 sample remains stable over this temperature range. The TG patterns of the composites are similar. They exhibit a fast weight loss at about 450–650°C, corresponding to the combustion of g-C3N4 and GO. After 800°C, the composites do not lose weight. At this point, CdMoO4/g-C3N4/GO changes to CdMoO4. So, the calculated CdMoO4 content is 7.19, 11.21, 23.47, and 33.38% for the M1, M2, M3, and M4 samples, respectively. In practice, these values are lower. Therefore, it is possible that only a part of Cd2+ and MoO42− ions adsorbs on the matrix and reacts with each other to form CdMoO4.
To evaluate the optical properties of the synthesized materials, we used UV-Vis-DRS. Fig. 3c presents the spectra of the CdMoO4, g-C3N4, GO, M2, M3, and M4 samples. It can be seen that CdMoO4 exhibits an absorption band in the ultraviolet region (<360 nm), corresponding to a band gap energy (Eg) of 3.44 eV (Fig. 3d). g-C3N4 has a band at a longer wavelength (>400 nm, Fig. 3c), corresponding to a band gap of 2.7 eV (Fig. 3d). The band gap values of CdMoO4 and g-C3N4 are similar to those reported by Bo Chai et al. [43]. For GO, the band gap energy depends on the synthesis method and conditions. Hunt et al. [62] reported a band gap between 1.6 and 2.1 eV; the values by Yeh et al. [63] were 2.3 and 2.7 eV, while Wang et al. [50] found no band gap. In this study, the spectrum of GO exhibits an absorption band in the visible light region, corresponding to a band gap of 2.43 eV, which is consistent with what was reported by Yeh et al. [63]. In our study, the composites and g-C3N4 have similar spectra. They absorb light in the visible region, with a slight blue shift (Fig. 3c). The calculated band gap energies are 2.75 (M2), 2.72 (M3), and 2.73 eV (M4) (Fig. 3d). These band gaps are slightly higher than that of g-C3N4 but much smaller than that of CdMoO4. Thus, the dispersion of CdMoO4 onto the g-C3N4 and GO matrices leads to a decrease in bandgap energy and an increase in visible light absorption. These findings show the photocatalytic activity potential of CdMoO4/g-C3N4/GO composites in the visible light region.
3.2. Photocatalytic Activities of The Synthesized Composites
The photocatalytic activity of the synthesized materials was evaluated via the decomposition of RhB under visible light. Fig. 4a presents the RhB degradation ability of the samples with different CdMoO4 contents. It can be seen that RhB degradation enhances with the CdMoO4 content from 7.2 to 23.5%, then diminishes at the higher CdMoO4 content (33.4%).
Photocatalytic degradation of RhB (a), first-order kinetics plot for the degradation of RhB (b) of composites with different CdMoO4 contents, and photocatalytic degradation of RhB of different materials (c), heavy metal leaching experiment (d). Reaction conditions: 125 mL RhB 10 mg/L + 25 mg catalyst, and 303 K.
The degradation efficiency of samples M1, M2, M3 and M4 is 55.7, 65.0, 89.1, and 82.2%, respectively, corresponding to rate constants of 0.0053, 0.0067, 0.0183, and 0.0139 min−1 (Fig. 4b). As stated previously, the composites’ surface area and band gap energy do not change significantly with increasing CdMoO4 content. Thus, the ability to degrade RhB depends mainly on the content of CdMoO4, with the highest value of 23.5%.
The RhB degradation ability of CdMoO4/g-C3N4/GO composites (sample M3), single materials (CdMoO4, g-C3N4, GO), and binary component composites (CdMoO4/GO (CMGO) or CdMoO4/g-C3N4 (CMCN)) was compared (Fig. 4c). The degradation of RhB over CdMoO4 and GO catalysts is negligible (15.5 and 13.3%). As is known, CdMoO4 is a semiconductor capable of operating in the ultraviolet region with a band gap of 3.44 eV and GO can slightly absorb visible light. Therefore, they do not show high activity under visible light. However, it can also be seen that because of the large specific surface area, GO has a higher ability to adsorb RhB than CdMoO4. g-C3N4 exhibits higher RhB degradation than CdMoO4 and GO because of its better absorption of visible light. The degradation efficiency of g-C3N4 is 40.5%. When dispersing CdMoO4 onto GO (CMGO sample) or g-C3N4 (CMCN sample), the resulting materials have an enhanced ability to decompose RhB with an efficiency of 54 and 65.3%. However, when CdMoO4 forms a ternary composite with GO and g-C3N4, the RhB degradation efficiency reaches 89%. As reported in previous studies [29,50,64], ternary component catalysts have better catalytic performance than their single or binary counterparts in the degradation of organic pollutants. According to our results, the combination of CdMoO4 with g-C3N4 and GO provides a composite with a band gap energy of about 2.7 eV, which is suitable for absorption in the visible region. Besides, this combination significantly increases the surface area of the composite. This increase, in turn, increases the heterojunction space between CdMoO4 and g-C3N4 and GO. As a result, this prevents the recombination of electrons and holes. In addition, the increase in the surface area of the composite also increases the adsorption capacity for RhB, thus enhancing the decomposition of RhB. The heavy metal leaching experiment was also processed where the catalyst was filtered after 120 min of reaction; the decolourization of RhB almost stopped despite further illumination (Fig. 4d). This result indicates that the CdMoO4/g-C3N4/GO composites is a heterogeneous catalyst in the photocatalytic degradation of RhB. Because heterogeneous catalysts usually release the metal ions to the solution, the presence of element Cd and Mo in the supernatant needs to be evaluated. It is found that the amount of Cd and Mo elements are too small to detect by using AAS. The effect of catalyst quantity on RhB degradation was also investigated. The catalyst quantity used in this experiment is 25–62.5 mg in 125 mL of 10 mg/L RhB solution (200–500 mg/L). Fig. S4a shows that the RhB degradation efficiency increases with the catalyst amount. After 120 min of illumination, the efficiency reaches 87.4, 96.6, 99.6, and 99.9% at the catalyst concentration of 200, 300, 400, and 500 mg/L, respectively. The decomposition rate calculated according to the pseudo-first-order kinetics equation is 0.0183, 0.0299, 0.0494, and 0.0616 min−1, respectively (Fig. S4b). The CdMoO4/g-C3N4/GO composite exhibits a photocatalytic activity for RhB degradation, equivalent to or higher than some previously published photocatalysts (Table 1).
Photocatalytic activity of some photocatalysts for RhB degradation
Temperature also affects the ability of the composite to degrade RhB, as shown in Fig. S5a. The rate of RhB decomposition increases rapidly with increasing reaction temperature. Decomposition efficiency reaches 76.05 and 99.55% at 293 and 303 K after 120 min of reaction, while at 313 K, the decomposition efficiency is 99.4% after 90 min. The decomposition rate constants calculated according to first-order kinetics at 293, 303, and 313 K are 0.0119, 0.0494, and 0.0673 min−1, respectively (Fig. S5b). Assuming the decomposition reaction rate follows the Arrhenius equation (Eq. (2)), we can calculate the activation energy from the slope of the line lnk versus 1/T (Fig. S5c). The obtained activation energy of 66.5 kJ/mol is relatively high, indicating that the decomposition of RhB strongly depends on temperature.
It is known that the photocatalytic degradation of dyes involves light absorption by the semiconductor to generate electrons (e−) and holes (h+). With a high oxidizing ability, holes can directly oxidize RhB molecules or react with H2O to generate (•OH) radicals. These free radicals oxidize the dye molecules adsorbed on the catalyst surface. Meanwhile, electrons combine with O2 molecules to create superoxide radicals (•O2−), and these radicals, in turn, oxidize RhB molecules [14,39,50]. In the case of CdMoO4/g-C3N4/GO, g-C3N4, with its low band gap (2.7 eV), absorbs visible light and generates electrons and holes. Then, the electrons move to the conduction band, and the holes move to the valence band of CdMoO4 because the conduction and valence bands of g-C3N4 are more negative than the conduction and valence bands of CdMoO4 [39,43,44]. GO in the composite increases the surface area of the catalyst. Furthermore, it increases the electron displacement, thereby reducing electron-hole recombination [49–51]. The photochemical degradation process can be described in Eq. (3 – 8) and the diagram in Fig. 5a.
Schematic depiction of the photocatalytic mechanism for RhB degradation by the synthesized composite (a) and the effect of radical scavengers on RhB degradation (b).
To obtain more information about the active species in the photocatalytic degradation of RhB, we conducted trapping experiments. The used radical scavengers are KI (quenching h+), IPA (quenching •OH), and BQ (quenching •O2−) in the same concentration of 5 mM [43,50,65]. Fig. 5b shows that the radical scavengers reduce the degradation of RhB. Therefore, it can be assumed that free radicals •O2−, •OH, and h+ are involved in the degradation of RhB, which is consistent with the mechanism described in the previous section. However, the influence of radical scavenging agents is different. With the RhB degradation efficiency after 120 minutes reaching 49.5, 29.6, and 23.5% in the presence of KI, IPA, and BQ, respectively, it can be concluded that BQ has a higher influence, followed by IPA and finally KI. In other words, the oxidant effects on RhB degradation in the order of •O2−> •OH > h+ hole.
The degradation of RhB was further investigated by using the HPLC-MS method. The HPLC results indicate that the photocatalytic reaction occurs very quickly. When the reaction time is equal to or longer than 30 min, the dye concentration in the solution remains too small to be determined with the MS method. Therefore, we took the samples after 15 min of reaction to identify their components. The results of MS spectra presented in Fig. S6 show that, after 15 min of reaction, the system has the following compounds: RhB (m/z = 443); N-(6-(diethylamino)-9-phenyl-3H-xanthen-3-ylidene)-N-ethylethanaminium (m/z = 399, (I)); N-(6-(ethyl(methyl)amino)-9-phenyl-3H-xanthen-3-ylidene) ethenaminium (m/z = 355, (II)); N-ethyl-N-(6-(ethyl(methyl)amino)-3H-xanthen-3-ylidene)ethanaminium (m/z = 311, (III)); N-(6-(ethylamino)-3H-xanthen-3-ylidene)-N-methylethana-minium (m/z = 282, (IV)) and 3-amino-N-ethyl-N-methyl-3H-xanthen-6-aminium (m/z = 255, (V)). According to previous reports [17,66], in the RhB photocatalytic degradation, oxidants such as •OH, •O2−, and h+ could attack the central carbon of RhB to cleave dye chromophore structure and produce low-weight intermediates that were further degraded by ring opening and mineralization. The MS analysis (Fig. S6) shows that intermediate (I) is the product of RhB decarboxylation, intermediate (II) is a product of intermediate (I) N-de-ethylation, intermediates (IV) and (V) are products of the N-de-ethylation of the intermediates in the previous step. The intermediate (III) is a product of both N-de-ethylation and cleavage of the C-C bond between the two benzene rings of intermediate (I). Thus, the photocatalytic degradation of RhB on the CdMoO4/g-C3N4/GO composite includes N-demethylation and dye chromophore structure cleavage. However, As mentioned above, because the concentration of other intermediates is too small when a longer reaction time, we cannot determine the existence of intermediates with smaller mass. Moreover, observing the COD change of the RhB degradation process showed that the COD decreased rapidly from 9.86 to 6.24, 4.48, 2.67, and 1.54 mg/L after reaction times of 15, 30, 60, and 90 min, respectively. Therefore, we suppose that the degradation of RhB on the CdMoO4/g-C3N4/GO composite could have been further carried out by the ring opening and mineralization, as previously reported [68–70].
The stability of the composite was evaluated via its reusability. Recycling experiments were performed under the same reaction conditions. After each cycle, the photocatalyst was filtered, washed with distilled water, and dried at 80°C. The results presented in Fig. 6a show that after five recycles, the photocatalytic activity of the composites still retain little changeably. The efficiency of RhB degradation is reduced about 10% after the five recycles. The decrease in photocatalytic activity of the composite is probably that a part of RhB has not been completely removed from the composite surface after each reuse. In addition, the XRD analysis of the composite after and five cycles was also performed (Fig. 6b). Almost no significant change was observed between the XRD samples before and after use. From these results, it can be assumed that the composite has high stability.
The RhB degradation performance of CdMoO4/g-C3N4/GO composite over five uses (a). XRD patterns of composite before and after five uses (b). Reaction conditions: 125 mL RhB 10 mg/L + 50 mg catalyst, and 303 K.
4. Conclusions
The CdMoO4/g-C3N4/GO ternary photocatalyst was successfully synthesized. CdMoO4 is highly dispersed on the g-C3N4 and GO sheets. The combination of CdMoO4, g-C3N4, and GO significantly enhances the RhB photocatalytic performance of the materials. The decomposition rate calculated according to pseudo-first-order kinetics is 0.0673 min−1, and the activation energy is 66.5 kJ/mol. The active specie •O2− plays the dominant role, followed by the •OH, and final is the h+ hole in RhB degradation. The degradation pathways include N-de-ethylation, chromophore cleavage, ring opening, and mineralization.
Supplementary Information
Acknowledgements
This research is funded by the University of Education, Hue University, Vietnam, under grant number T.21-TN.NCM-02.
Notes
Conflicts of interest
The authors declare that they have no conflicts of interest.
Author Contribution Statement
The authors confirm contribution to the paper as follows: study conception, design, draft and final manuscript preparation: H.V.D. (Associate Professor), D.N.N (Associate Professor) and D.Q.K. (Professor); data collection: N.T.A.T. (PhD) and D.T.N.C (Ms.); analysis and interpretation of results: N.D.L. (Associate Professor), N.T.A.N (Associate Professor) and N.L.M.L (PhD). All authors reviewed the results and approved the final version of the manuscript.