Introduction

The semiconductor photocatalysts have attracted much attention in several application fields including degradation of organic pollutants in water and production of hydrogen and/or oxygen via water splitting under light irradiation [1–5]. Among these semiconductor photocatalysts, TiO2 has been regarded as one of the most promising photocatalysts due to its high chemical stability, superior reusability, low cost, and low toxicity [6–8]. However, the wide band gap of anatase TiO2 (3.2 eV) leads it to low efficiency in utilization of visible light. Therefore, several approaches, such as metal or nonmetal doping [9, 10], noble metal deposition, and conjugated polymer modification [11, 12], have been developed to improve the visible-light photocatalytic activity of the TiO2 photocatalyst. Meanwhile, some new visible-light photocatalysts, such as silver halide [13], silver phosphate [14], sulfides [15], vanadates [16], BiXO (X = Br, I) [17], and graphic-like carbon nitride (g-C3N4) [18], have been prepared and investigated in detail.

Recently, the g-C3N4 photocatalyst has drawn a lot of interests in visible-light photocatalytic applications owing to its easy production and superior chemical stability [19, 20]. Nevertheless, the high recombination probability of photogenerated electron–hole pairs leads g-C3N4 to exhibiting low photocatalytic activity under visible-light irradiation, and g-C3N4 can be only excited by the visible light with the wavelength <450 nm due to its medium-band gap of ca. 2.7 eV, indicating that g-C3N4 cannot efficiently utilize the visible light with longer wavelength (>450 nm) to degrade organic pollutants and to produce hydrogen and oxygen via water splitting [21–25]. Therefore, many strategies have been developed to overcome the above-mentioned drawbacks, such as doping with metal or nonmetal elements, coupling with other semiconductors (e.g., WO3, CdS, AgX, graphene, reduced graphene oxide, TiO2, ZnO) [26–30], and modifying by conjugated polymers [30, 31].

Among these strategies, coupling g-C3N4 with CdS, a semiconductor with narrow band gap of 2.5 eV [32], has become a promising technique to prepare an efficient visible-light photocatalyst because of higher separation efficiency of photogenerated hole/electron pairs in CdS/g-C3N4 composite and efficient utilization of visible light with longer wavelength of 450 nm. Ge et al. [33] prepared CdS/g-C3N4 composite photocatalysts by coupling g-C3N4 powders with CdS quantum dots, Zhang et al. [34] prepared CdS/g-C3N4 core/shell nanowires using a combined solvothermal and chemisorption method, and Cao et al. [35] synthesized g-C3N4 nanosheets modified by CdS quantum dots through a solvothermal method. The CdS-modified g-C3N4 powders, nanowires, and nanosheets exhibited obviously improved hydrogen evolution rate via water splitting under visible-light irradiation. Fu et al. [36] prepared CdS/g-C3N4 composite photocatalysts via an “in situ” precipitation–deposition method, and Jiang et al. [37] fabricated CdS/g-C3N4 composite photocatalyst via a precipitation method. They found that the composite photocatalysts degraded organic compounds more efficiently than pure g-C3N4 under visible light. Dai et al. [38] investigated selective oxidation of aromatic alcohols to aromatic aldehydes and reduction of nitrobenzene into aniline using CdS/g-C3N4 composite as a photocatalyst under visible-light illumination. However, the above-mentioned investigations mainly dealt with CdS/g-C3N4 composites containing CdS more than 10 %, a larger amount of CdS in the composites can possibly cause a new environmental pollution due to the significant hazardousness of Cd2+. Therefore, it is a meaningful investigation to prepare an efficient CdS/g-C3N4 photocatalyst containing a small amount of CdS (≤1 %). So far, to the best of our knowledge, little work was investigated on the composite photocatalyst with the CdS content of <1 %.

In this work, the CdS/g-C3N4 composite photocatalysts were prepared via a surface-hybridization of g-C3N4 using a small amount of CdS nanoparticles. The as-prepared nanocomposites were characterized by X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), Raman spectroscopy, scanning electron microscopy (SEM), transmission electron microscopy (TEM), UV–Vis diffuse reflectance spectroscopy (DRS), photoluminescence spectroscopy (PL), electrochemical impedance spectroscopy (EIS), and photocurrent–time measurement (PT). The photocatalytic activity and stability were investigated by evaluating the photodegradation of Rhodamine B (RhB) under visible-light irradiation, and the main affecting factors of the visible-light photocatalytic activity were discussed.

Experimental

Reagents and materials

Cadmium acetate was purchased from Tianjin Guangfu Fine Chemical Research Institute. Sodium sulfide was obtained from Tianjin Bodi Chemical Company. Melamine and phenol were purchased from Aladdin Company. All the above reagents were of AR grade and used without further purification. All the solutions were prepared using deionized water.

Preparation of photocatalysts

The g-C3N4 samples were prepared by directly heating melamine in a semi-closed system according to the literature [39, 40]. In a typical preparation process, 20 g of melamine was firstly heated in a muffle furnace from room temperature to 520 °C with a heating rate of 2 °C min−1. After heating at 520 °C for 2 h, the as-prepared g-C3N4 was cooled naturally to room temperature, and then was grinded for further use.

The CdS/g-C3N4 composite photocatalysts were prepared according to the following steps. Firstly, a designed amount of cadmium acetate was dissolved in 50 g of deionized water. Secondly, 2.0 g of g-C3N4 powders was added into the cadmium acetate solution under ultrasonic conditions. After ultrasonification for 10 min, the obtained suspension was heated at 100 °C to obtain a dried mixture of g-C3N4 and cadmium acetate via solvent volatilization. Thirdly, the dried mixture was added into a sodium sulfide solution to form a CdS/g-C3N4 nanocomposite through a precipitation method. The as-prepared CdS/g-C3N4 nanocomposites with different CdS contents were labeled as CdS/g-C3N4(x %), where x % is the content of CdS in the CdS/g-C3N4 nanocomposites.

Characterization of photocatalysts

The XRD patterns of the CdS/g-C3N4 nanocomposites and pure g-C3N4 were measured in the range of 2θ = 10°–80° by continuous scanning on a Rigaku, D/MAX-2500 diffractometer (Rigaku Co., Japan) with Cu K α radiation (λ = 0.15406 nm). The operated voltage and current were 40 kV and 100 mA, respectively. XPS measurements were performed using a PHI 5000C ESCA system with Al Kα radiation (hν = 1486.6 eV) at a detection angle of 54°. The X-ray anode was run at 250 W, and the high voltage was kept at 15.0 kV. Raman spectroscopy was performed by a Nicolet 6700 Raman microspectrometer at a resolution of 2 cm−1, an Ar ion laser with the line of 514.5 nm was used as the excitation source, and the accelerating voltage and emission current were 30 kV and 30 mA, respectively. The emission SEM was performed using a HITACHI S-4800-I emission scanning electron microscope operated at accelerating voltage of 10 kV. The TEM measurement was performed using a JEM 2100 transmission electron microscope, and the operating voltage was 200 kV. UV–Vis DRS was performed on a SHIMADZU-2550 Scan UV–Vis system equipped with an integrating sphere attachment (Shimadzu Co., Japan) in the range of 200–800 nm, and BaSO4 was used as the background. The photoluminescence emission spectra (PL) of samples were detected by a Fluorescence spectrophotometer (F-4600 FL Spectrophotometer, Hitachi, Japan) at room temperature. The excitation source and light wavelength were a Xenon lamp and 316 nm, respectively. Electrochemical impedance spectra (EIS) were measured by an electrochemical system (Solartron 1255B frequency response analyzer and Solartron SI 1287 electrochemical interface) which was using 0.1 M KCl solution as the electrolyte, platinum electrode as counter electrode, saturated calomel electrode (SCE) as reference electrode, and FTO/CdS/g-C3N4 or FTO/g-C3N4 electrode as the working electrode. CdS/g-C3N4 and g-C3N4 films were coated by a doctor-blade method on the FTO substrates (fluorine-doped SnO2, 15Ω/sq). PT was measured on an electrochemical workstation (CHI660E, CHI Shanghai, Inc.), and 0.5 M Na2SO4 solution was used as the electrolyte, platinum electrode as counter electrode, Ag/AgCl electrode as reference electrode, and CdS/g-C3N4 or g-C3N4 electrode as the working electrode. The Brunauer–Emmett–Teller (BET) surface area measurements were performed using a Micromeritics TriStar II 3020 surface area and porosity system using nitrogen as adsorption gas at 77 K.

Photocatalytic activity measurement

The visible-light photocatalytic activity of the investigated materials was evaluated by RhB and phenol photodegradation. The photocatalytic degradation of RhB or phenol was carried out by decomposing RhB solutions (100 mL, 4 mg L−1) or phenol solution (100 mL, 20 mg L−1) containing 0.10 g of photocatalysts under visible-light irradiation. Before irradiation, the suspensions were continuously stirred for 1 h in the dark to establish adsorption/desorption equilibrium of RhB (or phenol) between the RhB (or phenol) solution and the photocatalyst surface. Afterward, a 300 W iodine tungsten lamp (Philips Co.) was used as a visible-light source with a 400 nm optical filter, and the luminous intensity on the liquid surface was ca. 85.0 mW cm−2 measured by an optical power meter (Model 842-PE, Newport Corporation, USA). During the irradiation period, the samples were taken out every 30 min or 1 h from the reactor, and then the photocatalysts were separated from the suspension by high-speed centrifugation. The clarified solution was analyzed by a T6 UV–Vis spectrometer (Beijing Purkinje General Instruments Co., Ltd., China), and the absorbance of RhB and phenol was measured at their maximum absorption wavelengths of 554 and 270 nm, respectively. The symbols of c 0 and c are the concentrations of the RhB or phenol solution before and after photoirradiation, respectively.

Results and discussion

Characterization of CdS/g-C3N4 nanocomposites

Figure 1a illustrates the XRD patterns of pure CdS, pure g-C3N4, and CdS/g-C3N4 nanocomposite photocatalysts. The two distinct diffraction peaks at 27.7° and 13.1° in the XRD spectrum of pure g-C3N4 can be indexed as the (002) and (100) planes, which come from the interlayer stacking of aromatic segments and in-plane structural packing motif, respectively. The three characteristic peaks at 52.06°, 43.98°, and 26.41° in the XRD pattern of pure CdS can be described to the (311), (220), and (111) planes of cubic CdS (JCPDS No.10-0454). From the XRD spectra of CdS/g-C3N4 nanocomposite photocatalysts, the two characteristic peaks of g-C3N4 are clearly observed, while the characteristic peaks of CdS are not observed because of its low content in the nanocomposites. The magnified XRD spectra of the nanocomposites (Fig. 1b) display a peak at 43.98°, which is ascribed to the characteristic one of cubic CdS. This result confirms the existence of CdS in the CdS/g-C3N4 nanocomposite photocatalysts.

Fig. 1
figure 1

XRD patterns of pure CdS, pure g-C3N4, and CdS/g-C3N4 nanocomposites (a), and magnified XRD patterns of CdS/g-C3N4 nanocomposites (b)

Full size image

XPS analysis is widely used to determine the chemical composition of the materials surface. The XPS survey spectra of the investigated samples are displayed in Fig. 2. The signals of C, N, O, Cd, and S can be observed from XPS spectrum of CdS/g-C3N4(1 %), while only the signals of C, N, and O can be observed from that of pure g-C3N4. The element of N originates from g-C3N4, the element of C is attributed to g-C3N4 and adventitious elemental carbon, both Cd and S are ascribed to CdS, and O comes from the environment. This result and EDS result shown in Fig. S1 further confirm that the CdS exactly exists on the g-C3N4 surface, accordant with the XRD result.

Fig. 2
figure 2

XPS survey spectra of pure g-C3N4 (a) and CdS/g-C3N4(1 %) (b)

Full size image

Figure 3 shows the Raman spectra of CdS/g-C3N4(1 %), pure g-C3N4, and pure CdS. No characteristic peaks of pure CdS can be observed in its Raman spectrum, and the Raman spectrum of CdS/g-C3N4(1 %) is same as that of pure g-C3N4, indicating that the existence of CdS does not change the chemical composition of g-C3N4.

Fig. 3
figure 3

Raman spectra of CdS/g-C3N4(1 %), pure g-C3N4, and pure CdS

Full size image

The morphology and microstructure of CdS/g-C3N4(1 %) and pure g-C3N4 were investigated by SEM and TEM, and the results are shown in Figs. 4 and 5, respectively. The SEM images in Fig. 4 show that both CdS/g-C3N4(1 %) and pure g-C3N4 display layered structure, and the morphology of the two samples is similar to each other except for the broken evidence of CdS/g-C3N4(1 %) due to its ultrosonification treatment. The TEM images in Fig. 5 further reveal the layered structure of pure g-C3N4 and CdS/g-C3N4(1 %). The CdS nanoparticles with the size of about 5 nm obviously aggregate and contact with g-C3N4 (Fig. 5b). The HRTEM image (set in Fig. 5b) of CdS nanoparticles exhibits fringe spacing of 0.337, 0.206, and 0.176 nm, which are ascribed to the (111), (220), and (311) lattice planes of cubic CdS (JSPDS-10-0454), respectively.

Fig. 4
figure 4

SEM images of pure g-C3N4 (a) and CdS/g-C3N4(1 %) (b)

Full size image
Fig. 5
figure 5

TEM images of pure g-C3N4 (a) and CdS/g-C3N4(1 %) (b)

Full size image

The UV–Vis diffuse reflectance spectra (DRS) of pure g-C3N4 and CdS/g-C3N4 nanocomposites are shown in Fig. 6. As can be seen clearly, the absorption of the CdS/g-C3N4 nanocomposites is obviously higher than that of pure g-C3N4 in the range of 200–550 nm, and increases with the CdS content increasing. The improved absorption in the range of 450–550 nm is advantageous to photocatalytic activity of CdS/g-C3N4 nanocomposites under visible-light irradiation.

Fig. 6
figure 6

UV–Vis diffuse reflectance spectra of pure g-C3N4 and CdS/g-C3N4 nanocomposites

Full size image

Figure 7 displays the PL spectra of pure g-C3N4 and the CdS/g-C3N4 nanocomposites. A characteristic peak of PL spectra can be easily observed at about 450 nm, which is ascribed to the recombination of photogenerated electrons and holes in semiconductors [41]. The peak intensity of CdS/g-C3N4 nanocomposites is significantly lower than that of pure g-C3N4, and further weakens as the CdS content increases. This result indicates that the separation efficiency of photogenerated electron–hole pairs in CdS/g-C3N4 nanocomposites is enhanced, and the larger content of CdS in the nanocomposites favors the improvement in separation efficiency of photogenerated electron–hole pairs.

Fig. 7
figure 7

PL spectra of pure g-C3N4 and the CdS/g-C3N4 nanocomposites

Full size image

The EIS is a useful tool to investigate the charge transfer rate at semiconductor/electrolyte interface in the dark [42]. The EIS Nyquist plots of FTO/g-C3N4 electrode and FTO/CdS/g-C3N4(1 %) electrode are shown in Fig. 8. Compared with FTO/g-C3N4 electrode, FTO/CdS/g-C3N4(1 %) electrode exhibits a smaller arc radius on its EIS Nyquist plot, revealing that the separation efficiency of charges in CdS/g-C3N4(1 %) can be increased. This result can be further confirmed by the photocurrent (I ph) measurement which displays greatly enhanced photocurrents of the nanocomposites over pure g-C3N4 (Fig. S2). The improved separation efficiency of charges in the nanocomposites is advantageous to their visible-light photocatalytic activity.

Fig. 8
figure 8

EIS of FTO/g-C3N4 and FTO/CdS/g-C3N4(1 %) electrodes with an applied bias potential of −0.5 V

Full size image

Visible-light photocatalytic activity of CdS/g-C3N4 nanocomposites

Figure 9a shows the RhB photodegradation under visible-light irradiation in the presence of pure g-C3N4 and CdS/g-C3N4 nanocomposites. The photodegradation rate of RhB in the presence of the nanocomposites is obviously higher than that of pure g-C3N4, indicating that the nanocomposites exhibit enhanced visible-light photocatalytic activity. The relationship between ln(c 0 /c) and irradiation time, shown in Fig. 9b, reveals that the kinetics of RhB photodecomposition on the surface of investigated photocatalysts can be described by the apparent first-order reaction [43]. The photodegradation rate constants k of RhB can be obtained from the slopes of the straight lines in Fig. 9b. The values of k reveal that the visible-light photocatalytic activity of CdS/g-C3N4(0.3 %) is 1.9 times of that of pure g-C3N4, indicating that a small amount of CdS nanoparticles can greatly improve the visible-light photocatalytic activity of g-C3N4 photocatalyst.

Fig. 9
figure 9

Temporal course of the decrement in the RhB concentration with irradiation time (a) and relationships between ln(c 0/c) and irradiation time (b). Experimental conditions: RhB concentration (4 mg L−1), catalyst concentration (1 g L−1)

Full size image

For eliminating the sensitization effect of RhB, phenol, a colorless and widely used in photocatalysis investigations, was chosen as a second model pollutant. The results of phenol photodegradation catalyzed by pure g-C3N4 and CdS/g-C3N4(1 %) under visible-light irradiation are presented in Fig. 10. As can be clearly seen from Fig. 10, the phenol photodegradation rate content (0.0724 h−1) in the presence of CdS/g-C3N4(1 %) is obviously higher than that (0.0386 h−1) of pure g-C3N4, further confirming that the nanocomposite exhibits enhanced visible-light photocatalytic activity.

Fig. 10
figure 10

Temporal course of the decrement in the phenol concentration with irradiation time (a) and relationships between ln(c 0/c) and irradiation time (b). Experimental conditions: Phenol concentration (20 mg L−1), catalyst concentration (1 g L−1)

Full size image

To further confirm RhB degradation photocatalyzed by CdS/g-C3N4(1 %) under visible-light irradiation, we have carried out the total organic carbon (TOC) measurement (TOC-VCPH, Shimadzu, Japan), and the result is shown in Fig. S3. The RhB solution (10 mg L−1) was nearly decolorized after 3 h under visible-light irradiation, and about 90 % decrease of TOC is observed after 12 h, indicating that RhB has been not only decolorized but also mineralized efficiently.

Besides the photocatalytic activity, the photocatalytic stability is another essential factor for a photocatalyst [44]. The photocatalytic stability of CdS/g-C3N4(1 %) was investigated by the recycling experiments. After every cycle, the suspension was filtered to regenerate the photocatalysts and reused in the next cycle. Figure 11 shows that the photodegradation degree of RhB decreases a little with the recycling runs increasing, indicating that CdS/g-C3N4(1 %) exhibits good visible-light photocatalytic stability.

Fig. 11
figure 11

Effect of cycling runs on RhB degradation in the presence of CdS/g-C3N4(1 %) under visible-light irradiation

Full size image

Affecting factors of RhB photodegradation

CdS content in CdS/g-C3N4 nanocomposites

Figure 9 displays the effect of CdS content in CdS/g-C3N4 nanocomposites on their visible-light photocatalytic activity. It can be clearly found that the visible-light photocatalytic activity of the nanocomposites significantly increases with the increment of the CdS content, revealing that the higher CdS content in the nanocomposites favors their visible-light photocatalytic activity.

Initial RhB concentration

Figure 12 shows the relationships between ln(c 0 /c) and irradiation time at different RhB concentrations, and the inset of Fig. 12 displays the effect of initial RhB concentrations on the RhB photodegradation rates. As can be clearly seen, the initial RhB photodegradation rate increases at first and then decreases, and reaches the maximum value when the initial RhB concentration is 4 mg L−1. Generally, the RhB photodegradation rate is possibly affected by both the adsorption amount of RhB on the photocatalyst surface and photo-adsorbed ability of the photocatalysts. The increasing adsorption of RhB and photo-adsorbed capacity are advantageous to the photocatalytic degradation of RhB. As the initial RhB concentration increases, the adsorbed amount of RhB on the photocatalyst’s surface increases, while the photo-adsorbed ability of the photocatalysts decreases because of the decrement in the transmittance of RhB solutions. When the initial RhB concentration increases from 2 to 4 mg L−1, the adsorption amount of RhB on the photocatalyst surface changes more significantly (from 526 to 1839 μg g−1) than the transmittance of RhB solutions (from 38 to 15 %) (Figs. S4 and S5), indicating that the adsorption amount of RhB on the photocatalyst surface is the main affecting factor for the RhB photodegradation, thus the initial RhB photodegradation rate increases. When the initial concentration of RhB is over 4 mg L−1, the adsorption of RhB on photocatalyst surface changes less (from 1839 to 2374 μg g−1) than the transmittance of RhB solutions (from 15 to 1 %) (Figs. S4 and S5), and the transmittance of the RhB becomes the main affecting factor, so the photocatalytic degradation rate of MO begins to decrease.

Fig. 12
figure 12

Influence of the initial RhB concentrations on the RhB photodegradation rates (The inset shows the initial photodegradation rate as a function of initial RhB concentration). Experimental conditions: Catalyst concentration (1 g L−1)

Full size image

pH value

Figure 13 presents the effect of pH value of the investigated systems on the RhB photodegradation rate under visible light, revealing that the visible-light photocatalytic activity of the CdS/g-C3N4 nanocomposites decreases with the pH value increasing. The result can be explained as follows. The strong interaction exists between g-C3N4 and RhB molecules due to –N= in g-C3N4 and –COOH in RhB (The RhB molecular structure can be seen in Fig. S6), leading to high adsorption of RhB on the g-C3N4 surface, which is advantageous to the visible-light photocatalytic activity of photocatalysts. As the pH value of the investigated systems increases, the groups of –COOH in RhB gradually change into –COO−, leading to the decreasing interaction between g-C3N4 and RhB. This decreasing interaction can reduce the adsorption of RhB on the photocatalyst surface (Fig. S7), which is disadvantageous to the visible-light photocatalytic activity of the CdS/g-C3N4 photocatalysts. Therefore, the RhB photodegradation rate under visible light can be obviously decreased by increasing the pH value of the instigated systems.

Fig. 13
figure 13

Influence of pH value on the RhB photodegradation rate (The inset shows the initial photodegradation rate as a function of pH value). Experimental conditions: catalyst concentration (1 g L−1), initial RhB concentration (4 mg L−1)

Full size image

Concentration of CdS/g-C3N4 nanocomposites

The effect of concentration of CdS/g-C3N4 nanocomposites on their visible-light photocatalytic activity is shown in Fig. 14. As the concentration of CdS/g-C3N4 nanocomposites increases, their visible-light photocatalytic degradation rates increase obviously due to the larger amount of active sites on the investigated photocatalysts.

Fig. 14
figure 14

Effect of the concentration of CdS/g-C3N4 nanocomposites on the RhB photodegradation rate. Experimental conditions: Initial RhB concentration (4 mg L−1)

Full size image

Photocatalytic mechanism of CdS/g-C3N4 nanocomposites

The energy gaps of CdS and g-C3N4 are well known to be ca. 2.5 and 2.7 eV, respectively, so both CdS and g-C3N4 can easily absorb visible light and be excited to produce photogenerated holes and electrons [45, 46]. The photogenerated electrons in conduction band (CB) of g-C3N4 and CdS can react with O2 to form •O2 − due to the negative CB potentials of g-C3N4 (−1.13 V) and CdS (−0.48 V), and the radicals of •O2 − can further react with H+ to form •OH [47, 48]. The photogenerated holes in valence band (VB) of both CdS and g-C3N4 can directly degrade RhB molecules [48]. The holes, •O2 −, and •OH can be responsible for the degradation of organic compounds.

The photogenerated electrons in CB of g-C3N4 (−1.13 V) can be easily injected into CB of CdS (−0.43), and the photogenerated holes in VB of CdS (2.02 V) can efficiently transfer to VB of g-C3N4 (1.57 eV). This means more efficient separation between the photogenerated holes and electrons in CdS/g-C3N4 nanocomposite photocatalysts, which is advantageous to their visible-light photocatalytic activity. Therefore, the visible-light photocatalytic activity of the CdS/g-C3N4 nanocomposites is much higher than that of pure g-C3N4. The proposed photocatalytic mechanism of the CdS/g-C3N4 nanocomposites is schematically shown in Fig. 15, and the major possible reactions for the RhB photodegradation are displayed as follows.

Fig. 15
figure 15

The proposed visible-light photocatalytic mechanism of CdS/g-C3N4 nanocomposites

Full size image
$$ {\text{CdS}}/{\text{g}} - {\text{C}}_{3} {\text{N}}_{4} + {h}\nu \, \to {\text{CdS}}*/{\text{g}} - {\text{C}}_{3} {\text{N}}_{4} * $$
$$ {\text{CdS}}*/{\text{g}} - {\text{C}}_{3} {\text{N}}_{4} * \, \to {\text{CdS}}({\text{e}}^{ - } {\text{e}}^{ - } /{\text{h}}^{ + } {\text{h}}^{ + } )/{\text{g}} - {\text{C}}_{3} {\text{N}}_{4} ({\text{e}}^{ - } {\text{e}}^{ - } /{\text{h}}^{ + } {\text{h}}^{ + } ) $$
$$ {\text{CdS}}({\text{e}}^{ - } {\text{e}}^{ - } /{\text{h}}^{ + } {\text{h}}^{ + } )/{\text{g}} - {\text{C}}_{3} {\text{N}}_{4} ({\text{e}}^{ - } {\text{e}}^{ - } /{\text{h}}^{ + } {\text{h}}^{ + } ) \to {\text{CdS}}({\text{e}}^{ - } {\text{e}}^{ - } {\text{e}}^{ - } /{\text{h}}^{ + } )/{\text{g}} - {\text{C}}_{3} {\text{N}}_{4} ({\text{e}}^{ - } /{\text{h}}^{ + } {\text{h}}^{ + } {\text{h}}^{ + } ) $$
$$ {\text{e}}^{ - } + {\text{O}}_{2} \to \bullet {\text{O}}_{2}^{ - } $$
$$ \bullet {\text{O}}_{2}^{ - } + {\text{H}}^{ + } \to \, \bullet {\text{OOH}} $$
$$ \bullet {\text{OOH}} + {\text{H}}^{ + } + {\text{e}}^{ - } \to {\text{H}}_{2} {\text{O}}_{2} $$
$$ {\text{H}}_{2} {\text{O}}_{2} + {\text{e}}^{ - } \to \, \bullet {\text{OH}} + {\text{OH}}^{ - } $$
$$ {\text{h}}^{ + } + {\text{RhB }} \to {\text{Degradation products}} $$
$$ \bullet {\text{OH}} + {\text{RhB}} \to {\text{Degradation products}} $$
$$ \bullet {\text{O}}_{2}^{ - } + {\text{RhB}} \to {\text{Degradation products}} $$

Conclusions

An efficient visible-light photocatalyst was successfully synthesized by surface-hybridization of g-C3N4 using a small amount of CdS nanoparticles. The surface-hybridization of CdS nanoparticles with the content of <1 % is able to obviously improve the absorbance of g-C3N4 particles in the visible light range and separation efficiency of photogenerated electron/hole pairs in the nanocomposites, while hardly affects their crystallinity and nanocrystal size. The visible-light photocatalytic activity of CdS/g-C3N4 nanocomposites is much higher than that of pure g-C3N4, and increases with the increment in both CdS content on the nanocomposites and photocatalyst concentration, while increases at first and then decreases with the increment in both initial RhB concentration and pH value of the investigated system. The cycled experiments reveal that the visible-light photocatalytic stability of the investigated nanocomposites is good. Therefore, it is a facile method to prepare an efficient visible-light photocatalyst by the surface-hybridization of g-C3N4 using a small amount of CdS nanoparticles.