Photochemical preparation of the ternary composite CdS/Au/g-C₃N₄ with enhanced visible light photocatalytic performance and its microstructure

Abstract

A ternary composite photocatalyst CdS/Au/g-C₃N₄ was synthesized by a facile two-step photoreduction method. In this hybrid structure, CdS nanoparticles and CdS@Au core–shell nanoparticles are homogeneously deposited on the surface of g-C₃N₄ to produce two different structures CdS–g-C₃N₄ and CdS@Au–g-C₃N₄. The microstructure analysis results of the composition, chemical states, and optical properties of as-prepared ternary hybrid reveal that a strong interaction exists between CdS nanoparticles, CdS@Au nanoparticles and the g-C₃N₄ bulk. The ternary composite CdS/Au/g-C₃N₄ exhibits significantly enhanced photocatalytic activity for RhB degradation under visible light irradiation compared with the binary composites CdS/g-C₃N₄, Au/g-C₃N₄ and pure g-C₃N₄. Meanwhile, CdS/Au/g-C₃N₄ also demonstrates good photostability in the recycling test of RhB degradation. The active species in the photocatalytic reaction over CdS/Au/g-C₃N₄ are detected as O₂˙⁻ and h⁺. The superior photocatalytic performance of CdS/Au/g-C₃N₄ should be attributed to the enhanced separation efficiency of photogenerated electrons–holes induced by the formed heterojunctions CdS–g-C₃N₄ and Z-scheme structure CdS@Au–g-C₃N₄ where Au nanoparticles serve as electrons mediator.

---

1. Introduction

In recent years, carbon nitride with a graphitic-like structure (g-C₃N₄) as a photocatalyst has been widely researched for water splitting,^1–3 CO₂ reduction^4,5 and environment remediation^6,7 due to its unique two-dimensional structure, excellent chemical stability and tunable electronic structure. As g-C₃N₄ is a polymer semiconductor with a bandgap of 2.7 eV, it can effectively absorb visible light photons. Sunlight can be used as energy source in g-C₃N₄-based photocatalysis, which makes it an ideal visible light driven photocatalyst candidate. However, the visible light photocatalytic activity of g-C₃N₄ is seriously limited by its low specific surface area and quick recombination of photogenerated electron–hole pairs. The way to improve its photocatalytic activity under visible light irradiation is quite needed.⁸

Several efforts have been developed to overcome this drawback of g-C₃N₄. Surface modification of noble metals, for instance Ag,^9,10 Pt,^11,12 and particularly Au,^13,14 to improve the photocatalytic activity of g-C₃N₄ is one of the frequently used methods. It is believed that noble metal nanoparticles can serve as traps and enhance the transfer efficiency of photogenerated charges.¹⁵ Furthermore, the inherent surface plasmon resonance (SPR) of noble metal nanoparticles can strengthen the visible light absorption of bulk g-C₃N₄ and simultaneously provide the thermal redox active centre during photocatalytic reaction.¹⁶ Interestingly, Au nanoparticles loaded g-C₃N₄ nanosheets were also prepared. It is worth noting that the specific surface area of g-C₃N₄ nanosheets is increased by ultrasonication-assisted liquid exfoliation of bulk g-C₃N₄. The synergistic role of high surface area of g-C₃N₄ nanosheets and photogenerated charges transfer efficiency due to Au nanoparticles surface modification makes the nanohybrids show superior photocatalytic performance for methyl orange degradation under visible light irradiation compared to bulk g-C₃N₄, g-C₃N₄ nanosheets, and Au/bulk g-C₃N₄ hybrids.¹⁷

Another frequently used method for improving the photocatalytic activity of g-C₃N₄ is coupling it with the other semiconductors, such as TiO₂,^18–20 ZnO,^21,22 Bi₂WO₆,^23–25 CdS^26,27 and so on. Among various semiconductors, CdS has received much attention because it is one of the most attractive visible light active semiconductor photocatalysts for its desired bandgap width and suitable band edge position. Till now, CdS nanoparticles,^28,29 CdS nanorods,^30,31 CdS nanowires,³² and CdS quantum dots³³ have been used to modify bulk g-C₃N₄ in order to improve the visible light photocatalytic activity of g-C₃N₄. Besides this, some special nanostructures, such as CdS@g-C₃N₄ core–shell nanorods,³⁴ CdS@g-C₃N₄ core–shell nanowires,³⁵ and CdS nanoparticles modified g-C₃N₄ nanosheets³⁶ were also prepared. All the CdS/g-C₃N₄ structures have shown activity due to the well-matched band structure and intimate contact interfaces between g-C₃N₄ and CdS. In addition, the photocorrosion of CdS can be effectively inhibited in CdS/g-C₃N₄ hybrids by forming interfacial transfer of photogenerated charges between g-C₃N₄ and CdS, which leads to effective charge separation on both parts and inhibits the photocorrosion of CdS.

Very recently, a sandwich-structured sulfur-doped CdS/Au/g-C₃N₄ ternary photocatalyst were prepared by bath deposition method.³⁷ The composite showed an enhanced photocatalytic activity for water splitting and dye degradation compared to the binary heterojunctions CdS/g-C₃N₄ and Au/g-C₃N₄. The Z-scheme charges transfer mechanism in which Au nanoparticles serve as electron transfer mediator was proposed to explain the experimental results. Besides this, Au@CdS core–shell structure was also deposited on the surface of g-C₃N₄ to prepare a novel heterojunction structured composite photocatalyst CdS/Au/g-C₃N₄. The photocatalytic activity of the as-prepared composite photocatalyst for hydrogen production is about 125.8 times higher than that of pure g-C₃N₄ under visible light irradiation.³⁸ From these reports, it is seen that preparing a ternary photocatalyst that takes advantage of both noble metal nanoparticles and semiconductor is an effective method for further improving the photocatalytic activity of g-C₃N₄. However, the effects of Au and CdS co-modification on the microstructure and photodegradation mechanism for pollutants over CdS/Au/g-C₃N₄ photocatalysts have not been discussed in detail. Furthermore, it is supposed that the total all-solid-state Z structure of CdS@Au–g-C₃N₄ can be constructed by the photoreduction or deposition method due to the affinity of Au surface atoms to sulfur atoms.³⁹ However, this kind of interaction role is very weak. The binary heterojunction CdS–g-C₃N₄ may still exist in the ternary composite, which in turn influences its photocatalytic mechanism. In this work, a ternary photocatalyst CdS/Au/g-C₃N₄ was synthesized via a facile two-step photoreduction method which was different from the previous literatures.^37,38 The microstructures of CdS/Au/g-C₃N₄ were characterized in detail. Both CdS–g-C₃N₄ heterojunction and Z-scheme structure CdS@Au–g-C₃N₄ were observed in the composite. Furthermore, the main reactive species for RhB photodegradation over CdS/Au/g-C₃N₄ were verified by using various scavengers. Based on these results, the photodegradation mechanism was proposed and discussed. This work is instructive for the future researches on the construction of Z-scheme structure.

2. Experimental

2.1 Materials

Carbamide, dicyandiamide, nitric acid, chloroauric acid, cadmium nitrate, sulfur precipitated, potassium iodide (KI) and isopropyl alcohol (IPA) were purchased from Sinopharm Chemical Reagent Co., Ltd (Shanghai city, P. R. China). All reagents in this work are AR grade and used without further purification. The purity of argon and nitrogen gas is 99.9%. Deionized water is used throughout this work.

2.2 Synthesis of photocatalysts

In this work, g-C₃N₄ was prepared by thermal polymerization of precursor. In a typical process, 14 g of carbamide powder and 6 g of dicyandiamide powder were mixed and put into the milling tanks in a planetary ball mill. Grinding balls were added in the mixtures. The milling speed and milling time were set as 360 rpm and 1 hour. The obtained mixtures after mechanical milling were grinded to powder. 5 g of this powder was put in a alumina crucible and heated in a muffle furnace at 550 °C for 4 h with a ramp of 2 °C min⁻¹. After cooling down to room temperature, the light yellow products were obtained and then washed in 0.1 mol L⁻¹ nitric acid solution by continuous stirring for 1 h in order to remove the impurities in g-C₃N₄ powder. Finally, the as-prepared g-C₃N₄ was washed several times with deionized water and then dried in an oven at 80 °C for 12 h.

Au/g-C₃N₄ composites were synthesized by photoreduction method.¹⁷ In detail, 1.28 mL of chloroauric acid with concentration of 0.01 mol L⁻¹ and 8 mL of methanol was added in 100 mL deionized water. 0.5 g of g-C₃N₄ powder was added to the above solution and then ultrasonic oscillation for 15 min and stirring in dark for 30 min to achieve adsorption balance. The obtained suspension was purged with argon for 30 min, and then irradiated with a 300 W Xe lamp for 1 h. The as-prepared Au/g-C₃N₄ was washed several times with deionized water and then dried in an oven at 60 °C for 12 h.

CdS/Au/g-C₃N₄ composites were also synthesized by photoreduction method.⁴⁰ Firstly, 0.5 g of Au/g-C₃N₄ powder was dispersed in a mixed solution consisting of 60 mL ethanol and 40 mL deionized water by ultrasonic oscillation for 15 min and stirring in dark for 30 min. After that, 26.8 mg of cadmium nitrate and 17.8 mg of S₈ were added to the above solution and stirred for another 30 min. The obtained suspension was purged with argon for 30 min and then irradiated under UV-Vis light for 3 h provided with a 300 W Xe lamp. The as-prepared CdS/Au/g-C₃N₄ was washed several times with deionized water and then dried in an oven at 60 °C for 12 h. CdS/g-C₃N₄ composites were also synthesized with the same method through using g-C₃N₄ instead of Au/g-C₃N₄ in the first step.

2.3 Characterization

X-ray diffraction patterns were collected using a XD-2 diffractometer with Cukα radiation at a scan rate of 2° min⁻¹. Morphology of photocatalysts was analyzed with a FEI Tecnai G20 transmission electron microscopy with an acceleration voltage of 200 kV. The elementary composition and information on binding energies of samples were measured with a VG Multilab 2000X X-ray photoelectron spectroscopy. All binding energies were referenced to the C1s peak of surface adventitious carbon at 284.6 eV. UV-Vis diffuse reflectance spectra were conducted to study the optical absorption properties of different photocatalysts with BaSO₄ as a reflectance standard using a U-3900 spectrophotometer. The surface functional groups and chemical bands of the samples were identified by a Nicolet NEXUS-6700 Fourier transform infrared spectrometer. Photoluminescence spectra were recorded on an F-7000 spectrophotometer in order to characterize the photoluminescence intensity of the samples with an excitation wavelength of 370 nm.

Photoelectrochemical measurements were carried out on an electrochemical analyzer (CHI600E) in a standard three-electrode system. Pt flake and Ag/AgCl were provided as the counter electrode and the reference electrode, respectively. A 300 W Xe lamp with a 400 nm UV cut-off filter was used as the light source. The electrolyte was 0.1 M Na₂SO₄ solution. The preparation process of working electrode was similar as the other literature.²⁷ Typically, 50 mg of samples were mixed with 150 mg of ethyl cellulose firstly, and then stirred in 3 mL ethanol solution to obtain the homogeneous slurry. The slurry was coated onto an indium-tin oxide glass (ITO glass) by doctor blade method. The as-prepared sample films with the similar thickness were dried at 120 °C for 1 h to get working electrodes.

2.4 Photocatalytic activity tests

The photocatalytic activity of the as-prepared samples was evaluated by the photodegradation of RhB under visible light irradiation. Externally illumination was supplied by a 300 W Xe lamp with a 400 nm UV cut-off filter. Before illumination, 50 mg of the photocatalyst was dispersed in 100 mL RhB solution with a concentration of 10 mg L⁻¹ by ultrasonic oscillation for 15 min and continuous stirring for 30 min in dark to achieve adsorption–desorption equilibrium. Stirring was kept during the photocatalytic reaction to maintain the photocatalyst homogeneously dispersed in solution. 4 mL of the suspensions were collected from the reactor at different irradiation time interval and centrifuged at 10 000 rpm for 5 min to remove photocatalyst completely. The concentration changes of RhB were analyzed by a 2102PC UV-Vis spectrometer.

Five repeated experiments were conducted using the same testing conditions above for stability test. After each reaction, the photocatalysts were filtered and washed several times with deionized water.

The reaction species detection process is similar to the photodegradation experiments. Various scavengers were introduced to the RhB dye solution prior to the illumination.

3. Results and discussion

Fig. 1 shows the XRD patterns of g-C₃N₄, Au/g-C₃N₄, CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples. Two distinct diffraction peaks of g-C₃N₄ at 27.4° and 13.1° are clearly observed in all samples. The intensity of both diffraction peaks decreases with the coupling of CdS in CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples. It may be attributed to the strong interaction between g-C₃N₄ and CdS, suggesting that CdS has been successfully deposited on the surface of g-C₃N₄.⁴¹ However, Au surface modification doesn't influence the peak intensity of g-C₃N₄, which can be seen in the XRD pattern of Au/g-C₃N₄ sample.⁴² Three additional diffraction peaks at 27.1°, 44.1° and 52.1° attributed to the (111), (220) and (311) crystal plane of CdS are found in the XRD patterns of CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples. The peaks at 38.2°, 44.3° and 64.6° corresponding to the (111), (200) and (220) crystal plane of Au are also seen in Au/g-C₃N₄ sample.

Fig. 1 XRD patterns of g-C₃N₄, Au/g-C₃N₄, CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples.

Fig. 2 shows the TEM images of the different samples. It is clearly seen that g-C₃N₄ is composed of nanosheets with a laminar structure, as shown in Fig. 2a. Fig. 2b and c reveal that Au nanoparticles with diameter in the range of 5–20 nm as black dots are homogeneously dispersed on the surface of g-C₃N₄ in Au/g-C₃N₄ sample. Fig. 2d and e show the TEM image of CdS/g-C₃N₄. CdS nanoparticles with approximately size of 20 nm are deposited on the surface of g-C₃N₄ as gray dots. Some CdS nanoparticles are aggregated together. Fig. 2f and g show the TEM images of CdS/Au/g-C₃N₄ sample. It can be observed that the size of CdS nanoparticles is smaller than that of CdS/g-C₃N₄. Moreover, the CdS nanoparticles are dispersed more uniform than that of binary composite CdS/g-C₃N₄. Besides the individual CdS nanoparticles directly contacted with g-C₃N₄, the CdS coated Au nanoparticles (CdS@Au) are also seen in the sample, as shown in Fig. 2g. Fig. 2h shows the HRTEM image of CdS/Au/g-C₃N₄ sample. The (111) crystal plane of CdS crystal with a crystal plane space of about 0.34 nm is observed. It can be concluded that there are two different heterojunctions CdS–g-C₃N₄ and CdS@Au–g-C₃N₄ formed in the ternary composite CdS/Au/g-C₃N₄, which may definitely affect its photocatalytic activity. It has been reported that S₈ molecules are inclined to adsorbing on the surface of Au nanoparticles due to the affinity of Au surface atoms to sulfur atoms.³⁹ After that, the sulfur atoms are reduced to S²⁻ ions by the collected electrons in Au nanoparticles during the photochemical preparation process of Au/CdS/g-C₃N₄. The as-obtained S²⁻ ions are bonded to Cd²⁺ to form CdS shells around Au cores. As a result, Z-scheme structure CdS@Au–g-C₃N₄ is constructed. However, CdS–g-C₃N₄ structure is also observed in this work, indicating the interaction between Au and CdS nanoparticles is weak. The formation of total Z-scheme structure CdS@Au–g-C₃N₄ may be very difficult using photoreduction method.

Fig. 2 TEM images of as-prepared samples: (a) g-C₃N₄; (b) and (c) Au/g-C₃N₄; (d) and (e) CdS/g-C₃N₄; (f) and (g) CdS/Au/g-C₃N₄; (h) HRTEM image of CdS/Au/g-C₃N₄.

XPS measurements were conducted to investigate the chemical states of C, N, O, Cd, S, and Au in the as-prepared samples. Fig. 3a shows the XPS survey scans of g-C₃N₄, CdS/g-C₃N₄, Au/g-C₃N₄ and CdS/Au/g-C₃N₄. The elements of C, N, O, Cd, S, and Au have been detected in these samples. High resolution XPS spectra of C1s are shown in Fig. 3b. Three peaks at 284.6 eV, 285.6 eV and 288.0 eV have been deconvoluted, respectively. The peak at 284.6 eV can be attributed to adventitious carbon adsorbed on the surface. The peak at 288.0 eV is related to the sp² hybridized carbon in N-containing aromatic rings (N–C=N) of g-C₃N₄, and the peak at 285.6 eV is mainly assigned to C=N from the defect of g-C₃N₄.¹¹ With the incorporation of CdS nanoparticles, the peak at 285.6 eV for C1s in CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples shifts to the higher binding energy. However, no obvious change is observed on the C1s electronic structure with the addition of Au nanoparticles.

Fig. 3 (a) XPS survey scans of g-C₃N₄, Au/g-C₃N₄, CdS/g-C₃N₄, and CdS/Au/g-C₃N₄; high resolution XPS scan of as-prepared samples: (b) C1s; (c) N1s; (d) Cd3d; (e) S2p; (f) Au4f.

Fig. 3c exhibits the high resolution XPS scan of N1s in g-C₃N₄, which can be deconvoluted into four peaks at 398.4 eV, 399.1 eV, 400.4 eV and 403.7 eV, respectively. The main peaks at 398.4 eV and 399.1 eV are assigned to the sp² hybridized aromatic N to carbon atoms in the triazine units (C–N=C) and the tertiary N bonded to carbon atoms (N–(C)₃), respectively. The peak at 400.4 eV belongs to the amino groups (C–N–H) from the defect of g-C₃N₄. The weak peak at 404.4 eV is attributed to π-excitations, which comes from charging effects or positive charge localization in the heterocycles.¹¹ Similarly, the peak intensity at 403.7 eV for N1s in g-C₃N₄ is enhanced and the peak position shifts to a high binding energy. Both the high resolution XPS spectra of C1s and N1s reveal that an electronic interaction occurs between g-C₃N₄ bulk and CdS nanoparticles, indicating the intimate contact interfaces are produced in the sample.

In Fig. 3d, the high resolution XPS spectra of Cd3d are observed at 404.6 eV and 411.4 eV, corresponding to Cd²⁺ ions in the CdS nanoparticles. In Fig. 3e, the XPS peaks at 161.0 eV and 162.4 eV are assigned to the S2p in the CdS nanoparticles.²⁷ The analysis results of Fig. 3d and e confirm the presence of CdS nanoparticles in CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples.

In Fig. 3f, the high resolution XPS peaks at 83.6 eV and 88.2 eV corresponding to the Au4f_5/2 and Au4f_7/2 in metallic gold nanoparticles have been observed, indicating Au nanoparticles was successfully deposited on the surface of g-C₃N₄.¹⁶ It is worth noting that the Au4f peak intensity of CdS/Au/g-C₃N₄ is weaker than that of Au/g-C₃N₄, which may be due to the existence of CdS shell coated on the surface of Au nanoparticles.³⁸

The UV-Vis diffuse reflectance spectra of as-prepared samples are shown in Fig. 4. For each sample, a strong absorption at wavelength around 460 nm, corresponding to the optical bandgap of g-C₃N₄ at 2.7 eV is observed. Compared with pure g-C₃N₄ sample, Au/g-C₃N₄ and CdS/Au/g-C₃N₄ samples show a more intensive absorption at 560 nm in the visible light region. This broad absorption peak may arise from the SPR effect of Au nanoparticles.¹³ Furthermore, the absorption intensity at 560 nm of ternary composite CdS/Au/g-C₃N₄ is weaker than that of Au/g-C₃N₄ sample, which may be ascribed to the wrapping of CdS shell on the surface of Au nanoparticles.³⁸ Owing to the surface modification of CdS nanoparticles, an additional absorption at 520 nm is also observed in CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples.³³

Fig. 4 UV-Vis absorption spectra of g-C₃N₄, Au/g-C₃N₄, CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples.

Fig. 5a depicts the FTIR spectra of g-C₃N₄, Au/g-C₃N₄, CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples. In all samples, the sharp peak around 810 cm⁻¹ can be assigned to the characteristic breathing mode of triazine units in g-C₃N₄. Compared with g-C₃N₄ and CdS/g-C₃N₄, the Au/g-C₃N₄ and CdS/Au/g-C₃N₄ samples show enhanced absorption intensity in this region, as shown in Fig. 5b. This may be ascribed to the overlapping of Au–O band at 818 cm⁻¹ with the CN heterocycles of triazine units in g-C₃N₄.¹⁶ Several strong bands from 1200 cm⁻¹ to 1700 cm⁻¹ can be ascribed to the typical stretching modes of CN heterocyclic compounds, and the broad bands at 3100–3300 cm⁻¹ correspond to the stretching vibration of NH₂ or N–H groups in g-C₃N₄. No characteristic peaks between 1050 cm⁻¹ and 1660 cm⁻¹ related to the Cd–S bond of CdS nanoparticles appear in the curves of CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples due to the low content of CdS.³⁰

Fig. 5 (a) FTIR spectra of g-C₃N₄, Au/g-C₃N₄, CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples and (b) the magnified spectra ranged from 750 cm⁻¹ to 850 cm⁻¹.

Photoluminescence measurements of as-prepared samples have been carried out at an excitation wavelength of 370 nm, as shown in Fig. 6. The pure g-C₃N₄ sample exhibits a strong emission peak in the region from 400 nm to 600 nm. In comparison, the fluorescence peak intensity of Au/g-C₃N₄ and CdS/g-C₃N₄ decreases clearly. Particularly, the ternary composite CdS/Au/g-C₃N₄ presents the most significant diminished fluorescence peak intensity among all samples. For semiconductors, the photogenerated holes and electrons can bind after the excitation by incident light, which results in partial energy transferring to the fluorescence. The decrease of emission peak intensity for CdS/Au/g-C₃N₄ sample indicates its low binding efficiency between electrons and holes, which may be ascribed to the inhibiting role of Au nanoparticles.³⁷

Fig. 6 Photoluminescence spectra of g-C₃N₄, Au/g-C₃N₄, CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples.

Fig. 7 shows the transient photocurrent response of g-C₃N₄, CdS/g-C₃N₄, Au/g-C₃N₄ and CdS/Au/g-C₃N₄ electrodes under intermittent visible light irradiation. It is observed that the pure g-C₃N₄ sample demonstrates a low photocurrent density and an evident photocurrent response delay. In contrast, the composite photocatalysts Au/g-C₃N₄, CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ exhibit the increased photocurrent intensity and a fast photocurrent response. Particularly, CdS/Au/g-C₃N₄ has the highest photocurrent density among all the samples. Because the generation of photocurrent is the result of photo-generated electrons transferring from electrodes to ITO glass, the enhanced photocurrent density implies that more effective photogenerated electrons separation and transferring are achieved when the CdS@Au–g-C₃N₄ and CdS–g-C₃N₄ heterojunctions are constructed in CdS/Au/g-C₃N₄.

Fig. 7 Transient photocurrent response of g-C₃N₄, Au/g-C₃N₄, CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ samples.

The photocatalytic activities of as-prepared samples were determined by the degradation of organic dye under visible light irradiation. RhB as a colored dye was chosen as the model in this work, and results are shown in Fig. 8. RhB photodegradation without photocatalyst was carried out, and result exhibits that RhB is rarely degraded with the absence of photocatalysts. When pure g-C₃N₄ is used as photocatalyst, RhB has been completely degraded in 30 min. In contrast, the degradation time is notably reduced to 20 min by using Au/g-C₃N₄, CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ photocatalysts, indicating that Au and CdS nanoparticles surface modification can effectively enhance the photocatalytic activity of pure g-C₃N₄. Among the three different composites, the slope of degradation curves using CdS/Au/g-C₃N₄ photocatalyst is the sharpest, indicating its highest photocatalytic activity. The RhB photodegradation efficiency in 10 min over CdS/Au/g-C₃N₄ photocatalyst reaches 90.1%, while it reduces to 84.1%, 77.5%, and 43.0% for CdS/g-C₃N₄, Au/g-C₃N₄ and pure g-C₃N₄ photocatalysts. This result reveals that there is a synergistic role of Au and CdS nanoparticles surface modification.

Fig. 8 Photodegradation curves of RhB over g-C₃N₄, Au/g-C₃N₄, CdS/g-C₃N₄ and CdS/Au/g-C₃N₄ photocatalysts.

The photocatalytic stability of ternary composite CdS/Au/g-C₃N₄ was characterized by a recycling test of RhB degradation under visible light irradiation. As shown in Fig. 9, no significant decrease of photocatalytic activity is observed after five repeated experiments, indicating that CdS/Au/g-C₃N₄ is stable during the photocatalytic reaction.

Fig. 9 Cycling test in photocatalytic degradation of RhB by CdS/Au/g-C₃N₄ under visible light irradiation.

During photocatalytic process, a large number of active species including O₂˙⁻, ˙OH, h⁺ are involved in the degradation procedure of pollutant molecules.⁴² To investigate the influence of different active species on the photocatalytic reaction, KI and IPA were adopted as scavengers to quench h⁺ and ˙OH, respectively. The concentration of KI and IPA was 1 mmol L⁻¹. As the active species O₂˙⁻ mainly originates from the oxygen that dissolved in solution, N₂ was continuously purged to exhaust oxygen in solution in order to study the effect of O₂˙⁻ on the photocatalytic activity of CdS/Au/g-C₃N₄. As shown in Fig. 10, the photocatalytic activity of CdS/Au/g-C₃N₄ is greatly suppressed by N₂ purging as O₂˙⁻ scavenger, indicating that O₂˙⁻ is the main active species in the RhB photodegradation process. Similarly, a significant decrease in the photocatalytic activity is also observed when KI is used as h⁺ scavenger, suggesting h⁺ is another important active species in the reaction. However, a slightly decrease in the photocatalytic activity is obtained by the addition of IPA as ˙OH scavenger. This result reveals that ˙OH active species are not the determinant factor in the photocatalytic reaction.

Fig. 10 Effect of scavengers on photocatalytic degradation of RhB over CdS/Au/g-C₃N₄.

On the basis of above experiments and discussion, the photocatalytic mechanism over CdS/Au/g-C₃N₄ is illustrated in Fig. 11. As the ternary composite CdS/Au/g-C₃N₄ prepared in this experiment mainly consists of CdS/g-C₃N₄ and CdS@Au/g-C₃N₄ structures, two different mechanisms are proposed.

Fig. 11 Proposed photocatalytic mechanism of CdS/Au/g-C₃N₄: (a) CdS–g-C₃N₄ heterojunction; (b) Z-scheme structure CdS@Au–g-C₃N₄.

As shown in Fig. 11a, both CdS and g-C₃N₄ can generate electrons and holes under visible light irradiation. The electrons in valence band are photoexcited to the conduction band and leave the holes in valence band. It is reported that the conduction band (−0.5 V) and valence band (+1.9 V) of CdS are lower than that of g-C₃N₄.^43,44 Therefore, the generated electrons on the conduction band (−1.1 V) of g-C₃N₄ transfer to the conduction band of CdS directly, while the holes in the valance band of CdS migrate to the valence band (+1.6 V) of g-C₃N₄.³⁶ The intimate interfaces between CdS nanoparticles and g-C₃N₄ bulk greatly promotes this charges transfer process.

As observed in the TEM images of CdS/Au/g-C₃N₄, CdS nanoparticles with small size are homogeneously dispersed on the surface of g-C₃N₄. The heterojunctions between CdS and g-C₃N₄ are formed, thus enhancing the separation efficiency of photogenerated electron–hole pairs. The electrons can be involved in the O₂ to O₂˙⁻ reduction process because the conduction edge potential of CdS (−0.5 V) is more negative than the standard redox potential for E^o_(O2/O2˙⁻) (−0.33 V vs. NHE). However, the valence edge potential of g-C₃N₄ (+1.6 V) is more negative than the standard redox potential E^o_(˙OH/H2O) (+2.68 V vs. NHE), H₂O/˙OH oxidation reaction cannot be proceeded in this process.¹³ The holes might be involved in the oxidation reaction of RhB directly, which is accordance with the experimental results of scavenger.

Fig. 11b shows the other possible photocatalytic mechanism over the ternary composite CdS/Au/g-C₃N₄, which has been reported in previous works.^37,38 In this experiment, the Au@CdS–g-C₃N₄ structure in the ternary composite has been confirmed by the high resolution TEM results. Under visible light irradiation, the electrons and holes transfer mode may be consistent with the Z-scheme mechanism, in which Au nanoparticles serve as the electrons mediator. The photogenerated electrons on the conduction band of CdS and the excited holes on the valance band of g-C₃N₄ may combine at Au nanoparticles and annihilate. Meanwhile, the photogenerated electrons with strong reducibility would be left on the conduction band of CdS, and the holes with strong oxidability would be left on the valence band of g-C₃N₄. These photogenerated electrons and holes would be participated in the photocatalytic process.^37,38 As the valence edge potential of CdS (+1.9 V) is more negative than the standard redox potential E^o_(˙OH/H2O) (+2.68 V vs. NHE), the active species ˙OH are also not involved in the photocatalytic reaction. However, the self-sensitized mechanism of RhB could not be excluded in this experiment, which may also be contributed to its de-coloration.

Compared with the binary composite CdS/g-C₃N₄ and Au/g-C₃N₄, the ternary composite CdS/Au/g-C₃N₄ exhibits the higher photocatalytic activity. This would be ascribed to two reasons: one is that the size of CdS nanoparticles in ternary composite is much smaller than the binary composite. Furthermore, the highly dispersion of CdS nanoparticles may form more heterojunctions in CdS/Au/g-C₃N₄ than that in CdS/g-C₃N₄ sample. Thus, the separation efficiency of photogenerated electron–hole pairs in CdS/Au/g-C₃N₄ is enhanced. It is also believed the high separation and transfer efficiency in photocatalyst may improve its photostability.³⁵ The other reason is that the Z-scheme structure in CdS/Au/g-C₃N₄ provides the higher redox ability of photogenerated charges than CdS/g-C₃N₄ and Au/g-C₃N₄ samples. Therefore, the ternary composite CdS/Au/g-C₃N₄ shows the high photocatalytic activity and stability in this experiment.

4. Conclusions

In summary, a facile two-step photo-reduction method was applied to synthesize ternary composite photocatalyst CdS/Au/g-C₃N₄ in this study. Both CdS–g-C₃N₄ and CdS@Au–g-C₃N₄ heterojunctions are produced and observed in the hybrid structure. Compared with the binary photocatalyst CdS/g-C₃N₄, the size of CdS nanoparticles in ternary composite is much smaller. Moreover, the CdS nanoparticles in ternary composite distribute more uniform on the surface of g-C₃N₄ bulk. The RhB degradation results exhibit that the photocatalytic activity of ternary composite is significantly enhanced relative to the binary photocatalysts CdS/g-C₃N₄, Au/g-C₃N₄ and pure g-C₃N₄. The enhancement of photocatalytic activity may be due to the intimate interfaces formed between CdS and g-C₃N₄, and the Z-scheme charges transfer mode, which consequently enhance the separation efficiency of photogenerated electrons–holes and strengthen the redox ability of the photoexcited charges. Meanwhile, O₂˙⁻ and h⁺ are the main active species in the photocatalytic reaction over CdS/Au/g-C₃N₄.

Acknowledgements

The authors gratefully acknowledge the financial support of Natural Science Foundation of Hubei Province of China (2013CFA085), the National Natural Science Foundation of China (51202064, 51472081), Research Foundation for Talented Scholars of Hubei University of Technology (BSQD12119), and Open Foundation of Hubei Provincial Key Laboratory of Green Materials for Light Industry ([2013]2-22).

Notes and references

1. Q. Xiang, J. Yu and M. Jaroniec, J. Phys. Chem. C, 2011, 115, 7355–7363.
2. S. Yang, Y. Gong, J. Zhang, L. Zhan, L. Ma, Z. Fang, R. Vajtai, X. Wang and P. M. Ajayan, Adv. Mater., 2013, 25, 2452–2456.
3. X. Fan, L. Zhang, M. Wang, W. Huang, Y. Zhou, M. Li, R. Cheng and J. Shi, Appl. Catal., B, 2016, 182, 68–73.
4. J. Mao, T. Peng, X. Zhang, K. Li, L. Ye and L. Zan, Catal. Sci. Technol., 2013, 3, 1253–1260.
5. Q. Su, J. Sun, J. Wang, Z. Yang, W. Cheng and S. Zhang, Catal. Sci. Technol., 2014, 4, 1556–1562.
6. F. Liang and Y. Zhu, Appl. Catal., B, 2016, 180, 324–329.
7. Y. Cui, Z. Ding, P. Liu, M. Antonietti, X. Fu and X. Wang, Phys. Chem. Chem. Phys., 2012, 14, 1455–1462.
8. S. Cao, J. Low, J. Yu and M. Jaroniec, Adv. Mater., 2015, 27, 2150–2176.
9. L. Ge, C. Han, J. Liu and Y. Li, Appl. Catal., A, 2011, 409, 215–222.
10. Y. Yang, Y. Guo, F. Liu, X. Yuan, Y. Guo, S. Zhang, W. Guo and M. Huo, Appl. Catal., B, 2013, 142, 828–837.
11. W. J. Ong, L. L. Tan, S. P. Chai and S. T. Yong, Dalton Trans., 2015, 44, 1249–1257.
12. J. Yu, K. Wang, W. Xiao and B. Cheng, Phys. Chem. Chem. Phys., 2014, 16, 11492–11501.
13. J. Xue, S. Ma, Y. Zhou and Q. Wang, RSC Adv., 2015, 5, 88249–88257.
14. R. C. Pawar, S. Kang, S. H. Ahn and C. S. Lee, RSC Adv., 2015, 5, 24281–24292.
15. S. Chang, A. Xie, S. Chen and J. Xiang, Electroanal. Chem., 2014, 719, 86–91.
16. S. Samanta, S. Martha and K. Parida, ChemCatChem, 2014, 6, 1453–1462.
17. N. Cheng, J. Tian, Q. Liu, C. Ge, A. H. Qusti, A. M. Asiri, A. O. Al-Youbi and X. Sun, ACS Appl. Mater. Interfaces, 2013, 5, 6815–6819.
18. X. Song, Y. Hu, M. Zheng and C. Wei, Appl. Catal., B, 2016, 182, 587–597.
19. Y. Li, J. Wang, Y. Yang, Y. Zhang, D. He, Q. An and G. Cao, J. Hazard. Mater., 2015, 292, 79–89.
20. D. Lu, G. Zhang and Z. Wan, Appl. Surf. Sci., 2015, 358, 223–230.
21. S. P. Adhikari, H. R. Pant, H. J. Kim, C. H. Park and C. S. Kim, Ceram. Interfaces, 2015, 41, 12923–12929.
22. Y. He, Y. Wang, L. Zhang, B. Teng and M. Fan, Appl. Catal., B, 2015, 168, 1–8.
23. M. Li, L. Zhang, X. Fan, Y. Zhou, M. Wu and J. Shi, J. Mater. Chem. A, 2015, 3, 5189–5196.
24. L. Ge, C. Han and J. Liu, Appl. Catal., B, 2011, 108, 100–107.
25. H. Wang, J. Lu, F. Wang, W. Wei, Y. Chang and S. Dong, Ceram. Interfaces, 2014, 40, 9077–9086.
26. Y. Cui, Chin. J. Catal., 2015, 36, 372–379.
27. M. Lu, Z. Pei, S. Weng, W. Feng, Z. Fang, Z. Zheng, M. Huang and P. Liu, Phys. Chem. Chem. Phys., 2014, 16, 21280–21288.
28. F. Jiang, T. Yan, H. Chen, A. Sun, C. Xu and X. Wang, Appl. Surf. Sci., 2014, 295, 164–172.
29. J. Fu, B. Chang, Y. Tian, F. Xi and X. Dong, J. Mater. Chem. A, 2013, 1, 3083–3090.
30. J. Yuan, J. Wen, Y. Zhong, X. Li, Y. Fang, S. Zhang and W. Liu, J. Mater. Chem. A, 2015, 3, 18244–18255.
31. Z. L. Fang, H. F. Rong, L. Y. Zhou and P. Qi, J. Mater. Sci., 2015, 50, 3057–3064.
32. Y. Xu and W. D. Zhang, Eur. J. Inorg. Chem., 2015, 2015, 1744–1751.
33. L. Ge, F. Zuo, J. Liu, Q. Ma, C. Wang, D. Sun, L. Bartels and P. Feng, J. Phys. Chem. C, 2012, 116, 13708–13714.
34. Y. Li, X. Wei, H. Li, R. Wang, J. Feng, H. Yun and A. Zhou, RSC Adv., 2015, 5, 14074–14080.
35. J. Zhang, Y. Wang, J. Jin, J. Zhang, Z. Lin, F. Huang and J. Yu, ACS Appl. Mater. Interfaces, 2013, 5, 10317–10324.
36. S. W. Cao, Y. P. Yuan, J. Fang, M. M. Shahjamali, F. Y. Boey, J. Barber, S. C. J. Loo and C. Xue, Int. J. Hydrogen Energy, 2013, 38, 1258–1266.
37. W. Li, C. Feng, S. Dai, J. Yue, F. Hua and H. Hou, Appl. Catal., B, 2015, 168, 465–471.
38. X. Ding, Y. Li, J. Zhao, Y. Zhu, Y. Li, W. Deng and C. Wang, APL Mater., 2015, 3, 104410.
39. P. Zhou, J. G. Yu and M. Jaroniec, Adv. Mater., 2014, 26, 4920–4935.
40. H. Tada, T. Mitsui, T. Kiyonaga, T. Akita and K. Tanaka, Nat. Mater., 2006, 5, 782–786.
41. X. Dai, M. Xie, S. Meng, X. Fu and S. Chen, Appl. Catal., B, 2014, 158, 382–390.
42. R. C. Pawar, Y. Pyo, S. H. Ahn and C. S. Lee, Appl. Catal., B, 2015, 176, 654–666.
43. J. X. Sun, Y. P. Yuan, L. G. Qiu, X. Jiang, A. J. Shen and J. F. Zhu, Dalton Trans., 2012, 41, 6756–6763.
44. A. Sobczynski, A. J. Bard, A. Campion, M. A. Fox, T. Mallouk, S. E. Webber and J. M. White, J. Phys. Chem., 1987, 91, 3316–3320.

---