1 Introduction

The rapid development of modern industry and population increase have resulted in water contamination, which is gaining more and more attention worldwide. Organic dyes have become as one of the most significant water contaminants [1]. Many methods have been used to remove organic dyes from wastewater [2]. Among them, photocatalytic degradation of semiconductors is regarded as a promising and sustainable technology due to its ease of operation, high efficiency, economy, environmental friendliness, and good recyclability [3]. For example, Sheetal et al. fabricated a novel photocatalyst composed of CQDs-decorated MnIn2S4/Cu2O/Ag2O for the removal of MO. The degradation efficiency of MO displayed 98.6% of MO degradation in 70 min in visible light radiation [4]. The dye degradation reaction occurs when sunlight at the suitable frequency is combined with photoreactive compounds that absorb a broad range of the solar spectrum [5]. Under illuminated conditions, the valence band electrons of photocatalysts (such as TiO2 and ZnO) are excited to the conduction band, generating electron-hole pairs. Subsequently, these electrons and holes participate in redox reactions. Electrons can reduce dye molecules, while holes can oxidize water molecules and oxygen to produce hydroxyl radicals (·OH), superoxide radicals (·O2−), and other reactive oxygen species (ROS). ROS, potent oxidizing agents, can react with dye molecules, leading to the degradation of the dye’s chemical structure [6, 7].

So far, many inorganic photocatalysts, such as TiO2 [8], ZnO [9], and SnO2 [10], have been extensively researched for organic dye removal. Among them, ZnO is one of the most widely used semiconductors for photocatalysis because of its remarkable properties such as low toxicity, low cost, and high oxidizing capacity [11, 12]. Unfortunately, owing to its wide band gap (3.37 eV) and low electron-hole pairs separation efficiency, direct application of ZnO is limited [13]. Building a heterojunctions with other semiconductors with a narrower band gap has been reported as a desirable technique to overcoming the aforementioned ZnO-related disadvantages [14]. As it is well known, a band gap below 2.65–2.80 eV could absorb the maximum portion of the visible light spectrum.

Organic semiconductors have significant advantages over inorganic semiconductors due to their low toxicity, inexpensive cost, and plentiful reserves. g-C3N5, a nitrogen-rich graphitic carbon nitride, was successfully created with a greater nitrogen content (62.5 at%) than g-C3N4 (57.1 at%) [15]. In addition, g-C3N5 has a larger conjugated system than g-C3N4, attributed to the conjugation of the heptazine moiety to the lone electron pairs on the N atom [16, 17]. Accordingly, g-C3N5 possesses a narrower band gap (≈2.0 eV) than g-C3N4 (≈2.7 eV) [18, 19], which facilitates the improvement of solar energy utilization and photoactivity. Its wide application, however, is restricted by the smaller specific surface area and quick charge complexation. Doping with rare earth elements is thought to be an effective way to improve the light usage and energy conversion of g-C3N5. Rare earth metals have a unique valence electron structure that precludes photoelectron/hole complexation [20,21,22]. Among the rare earth metals, cerium (Ce) has better electrical conductivity, and the mixed valence states of 3-valence and 4-valence makes Ce-based materials attractive in heterogeneous catalysis [23, 24], so Ce-dopped g-C3N5 leads to different optical properties, which in turn leads to a high charge mobility to improve photocatalytic efficiency.

In summary, the key features of an excellent photocatalyst are a robust redox ability, high charge-separation efficiency, broad absorption range, and long-term stability. However, all of these features are unattainable in a single photocatalyst. As a result, organic and inorganic semiconductor composited photocatalyst is commonly employed to improve photocatalytic degradation of contaminates [25, 26]. However, photogenerated carrier’s redox ability is limited by the use of conventional heterojunctions [27]. In contrast, the newly developed S-scheme heterojunctions, which improve photocatalytic effectiveness by facilitating the charge transfer in nanocomposites, can retain the photogenerated charges’ high redox capacity while widening the light capture range [28, 29]. Jabbar et al. demonstrated that the S-scheme electron transfer mode between BiOCl and CuBi2O4 significantly enhanced the photooxidation ability of sulfamethoxazole over BiOCl/CuBi2O4 composites [30]. As a result, it is critical to develop new S-scheme heterojunctions photocatalytic materials capable of efficiently using photon energy in the visible range.

In this thesis, a novel ZN/Ce–CN heterojunctions photocatalysts with different cerium doping content were synthesized for the first time by hydrothermal and calcination methods for the degradation of MB pollutants. Compared to pure g-C3N5 and ZnO, ZN/Ce–CN, which featured a unique nanoparticle/flake coupling configuration. The photocatalytic degradation efficiency of ZN/Ce–CN for MB typically exceeds the efficiency of comparable published composite photocatalytic materials (Table 1). The boosted photocatalytic activity of ZN/Ce–CN was attributed to wide light response range and the improved charge separation resulting from the synergistic effect of the construction of the S-scheme heterojunctions between Ce–CN and ZnO. The ZN/Ce–CN developed in this study opens a new avenue for the development of a unique S-scheme heterojunction photocatalyst for photo- and electrocatalytic materials with a wide range of applications.

Table 1 The comparison of photocatalytic efficiency between published relevant studies and the current work for MB degradation is presented
Full size table

2 Experimental

2.1 Materials

Analytical grade zinc nitrate hexahydrate (Zn(NO3)2·6H2O), cerium nitrate hexahydrate (Ce(NO3)3·6H2O), 3-amino-1,2,4-triazole (3-AT), ethanol, isopropanol (IPA), methylene blue (MB), benzoquinone (BQ), sodium hydroxide (NaOH), concentrated sulfuric acid (H2SO4), potassium iodide (KI).

2.2 Synthesis of g-C3N5

To prepare the g-C3N5, 1.5 g of 3-AT was dissolved in 30 mL of distilled water and stirred at 80 °C on a magnetically heated stirrer until dry and crystallized, then ground into a powder and placed in a crucible for calcination in a muffle furnace at 5 °C min−1 up to 520 °C for 180 min.

2.3 Synthesis of Ce doped g-C3N5

The Ce-doped g-C3N5 is prepared in a similar manner to g-C3N5, except that Ce(NO3)3·6H2O with mass fractions of 0.25%, 0.5%, 1%, 2%, and 4% is dissolved before the addition of 1.5 g of 3-AT. The remaining steps are the same. The resulting sample is referred to as Ce–CN.

2.4 Preparation of ZnO

The first is the preparation of solution A, 2.97 g of Zn(NO3)2·6H2O was completely dissolved in 30 mL of distilled water, and anhydrous ethanol (20 mL) was added. Subsequently, 0.8 g of NaOH was dissolved in 15 mL of distilled water and progressively added dropwise to the aforementioned liquid A. The resultant solution was agitated for 30 min at room temperature, sonicated for 30 min, and then hydrothermally reacted at 100 °C for 10 h. The substance was then washed, dried, and processed to produce a fine white powder ZnO.

2.5 Preparation of ZN/Ce–CN

0.1 g of ZnO was added to 30 mL of anhydrous ethanol and agitated for 10 min at room temperature to obtain a milky white suspension. A brownish-yellow suspension was obtained by adding 0.1 g of Ce–CN to the solution, followed by stirring for 60 min. Next, the suspension underwent a 10-h hydrothermal reaction at 140 °C. After washing, drying, and grinding, a solid powder that is light brownish yellow is produced. The resulting sample is referred to as ZN/Ce–CN. It is sufficient to substitute g-C3N5 for the Ce–CN powder in the synthetic process for ZN/CN, and all other steps are the same. The preparation scheme of ZN/Ce–CN is shown in Scheme 1.

Scheme 1
scheme 1

Preparation scheme of ZnO/Ce-g-C3N5 (ZN/Ce–CN) composite

Full size image

2.6 Photocatalytic degradation of MB

The degradation trials were conducted at a distance of 5 cm from a xenon lamp (500 W), and the reaction temperature was maintained at 20 °C by a water recirculation system. To achieve the adsorption-desorption equilibrium in the dark, 30 mg of catalyst was added to a 40 mL MB solution (20 mg L−1, pH = 7) and agitated for 30 min. The photocatalytic reaction started when the xenon lamp was turned on and 3.0 mL of the solution was sampled periodically throughout the degradation process. After centrifugation to remove the catalyst, the concentration of MB supernatant was determined in a UV–Vis spectrophotometer (Labtech 9100B) based on the absorption peak at 665 nm.

2.7 Characterization

The phase structure of all samples were determined by X-ray diffraction (XRD). Fourier transform infrared (FT-IR) spectra were measured on a PerkinElmer infrared spectrometer Frontie. Transmission microscopy (TEM; JEOL, JEM-2100) was used to examine the structure and morphology of all samples. The UV–Vis diffuse reflectance spectra of the materials were acquired utilizing a Hitachi U-4100 UV–Vis spectrometer (Shimadzu, UV-2600). X-ray photoelectron spectroscopy (XPS, Axis Ultra, Kratos Analytical Ltd, UK) was used to analyze the prepared samples’ elemental composition and bonding information. The photoluminescence (PL) intensity of the prepared samples was characterized by fluorescence spectroscopy (PL). Electron paramagnetic resonance measurements were performed on a Bruker ER200-SRC.

3 Result and discussion

3.1 XRD patterns of different samples

XRD was used to analyze the crystalline characteristics and the phase structure of g-C3N5, ZnO, Ce–CN, ZN/CN, and ZN/Ce–CN (Fig. 1a, b). The diffraction peak of CN at 2θ = 27.4° was generated by the stacking of conjugated aromatic systems, corresponding to the (002) plane (JCPDS 87-1526). No peak corresponding to Ce species was observed in Ce–CN due to the trace quantity and homogeneous dispersion of Ce single atoms [31]. However, for Ce–CN catalyst, obvious shifts toward a higher 2θ value (about 0.2) were observed. This could be due to Ce doping, which induces distortion in the g-C3N5 lattice, similar to Ce-C3N4 [32]. All peaks in the range of 2θ = 30 ~ 80° were indexed according to the hexagonal wurtzite ZnO structure (JCPDS No. 99-0111) [33]. ZnO remains in its crystalline phase even after g-C3N5 hybridization. As a result, the diffraction peaks of ZN/Ce–CN contain each separate peak for Ce–CN and ZnO with different intensities. Furthermore, the absence of any other new peaks shows that the effective synthesis of ZN/Ce–CN was accomplished.

Fig. 1
figure 1

XRD patterns (a) and locally enlarged view of diffraction peaks (b) of g-C3N5, ZnO, Ce–CN, ZN/CN, and ZN/Ce–CN

Full size image

3.2 FTIR

FTIR spectra were also utilized to explore the characteristics of the as-prepared ZN/Ce–CN at each step (Fig. 2). Similar characteristic peaks are detected for typical CN networks in graphitic of virgin g-C3N5 and Ce–CN materials. The IR bands at 1240 cm−1, 1318 cm−1, 1411 cm−1, 1560 cm−1, and 1640 cm−1 are ascribed to the C–N stretch (νC–N) of heptazine (C6N7) aromatic nucleus in g-C3N5 as similar to g-C3N4 material. Whilst the characteristic broad peaks at 3154 cm−1 is attributed to symmetric and asymmetric stretching vibrations of the –NH2 (νN–H) and –OH (νO–H) groups [34]. The sharp peak observed at 812 cm−1 was related to the breathing mode of triazine units in g-C3N5 [35], and it was obviously found in the ZN/Ce–CN composite. There are no further vibration modes found for Ce–CN, indicating that the structure of g-C3N5 is unaffected by cerium doping. The entire graphite-like structure of cerium-doped is essential for the π-delocalized electron system, which aids in the generation and transportation of photogenerated carriers [36]. O–H stretching vibration accounts for the intense broadband at about 3450 cm−1 in the FTIR spectra of undoped ZnO [37]. The loading of ZnO broadens the peak bands of the O–H stretching vibration in ZN/CN and ZN/Ce–CN significantly. A well-defined band appears at 599–400 cm−1, which could be due to the stretching vibration of the Zn–O bond [14], and it was clearly identified with a little shift in the ZN/CN and ZN/Ce–CN composite. Notably, the successful synthesis of heterojunctions in ZN/Ce–CN can be confirmed by the characteristic peaks of ZN/Ce–CN containing both Ce–CN and ZnO.

Fig. 2
figure 2

FT-IR spectra of pure g-C3N5, ZnO, Ce–CN, ZN/CN and ZN/Ce–CN

Full size image

3.3 Morphological properties

TEM and HRTEM measurements were done to better understand the structure and content of the produced samples. It is evident from the ZN/Ce–CN TEM images (Fig. 3a) that ZnO nanoparticles were successfully deposited on Ce–CN nanosheets, demonstrating the composite of ZnO and Ce–CN, which is a fundamental requirement for heterojunctions production. Additionally, using HRTEM analysis, the interaction between the nanocomposites was investigated; the outcomes are displayed in Fig. 3b. HRTEM images further confirmed the presence of heterojunctions, with inter-planar distances of 0.281 and 0.247 nm corresponding to the (100) plane and (101) plane of ZnO, respectively [38]. Because of the low crystallinity, the lattice streaks of Ce–CN were not detectable. In addition, the utilization of energy dispersive spectroscopy (EDS) verified the existence of Zn, Ce, C, N, and O (Fig. 3d–i). The homogeneous distribution of all the elements suggests that the ZN/Ce–CN structure was successfully constructed.

Fig. 3
figure 3

TEM image (a), HRTEM image (b), FESEM figure (c), the corresponding elemental mapping images (d–i) of ZN/Ce–CN

Full size image

3.4 XPS analysis

The ZN/Ce–CN nanocomposite’s surface chemical composition and valence states were ascertained by XPS analysis; the outcomes are shown in Fig. 4. X-ray photoelectron spectroscopy (XPS) survey spectrum was utilized to confirm the presence of elements in ZN/Ce–CN nanocomposite which include C, O, N, Ce, and Zn, as shown in Fig. 4a, which coincides with the EDX analysis (Fig. 4i). In the HR-XPS spectrum of C 1s (Fig. 4b), the peaks at 284.9 eV and 287.9 eV which represent the C–C (sp3) and N–C=N (sp2) bonding, respectively [39]. The sp3 carbon peak was caused by the presence of residual unreacted precursor, turbostratic carbon, and adventitious carbon, whereas the sp2 carbon peak was caused by aromatic carbons building heptazine units (C6N7) of g-C3N5 matrix. The peak positioned at 398.5 eV in the HR-XPS spectra of N 1s (Fig. 4c) was assigned to the additive contribution of the secondary C=N–C/anionic nitrogen in the O–Zn–N linkage [23]. The typical peak at the binding energies of 399.4 eV corresponds to the tertiary (N–(C)3) nitrogen of the heptazine (C3N4) moiety. As a result of the cumulative contribution of bridging azo (C–N=N–C) nitrogen and residual –NH2 on the edge, a peak with a band energy value of 400.4 eV was detected. HR-XPS spectra of ZN/Ce–CN in the O 1s region (Fig. 4d) revealed two unique deconvoluted peaks at BE ≈ 530.6 and 531.7 eV. The former corresponds to the lattice oxygen of ZnO, whereas the latter corresponds to surface-adsorbed OH species. The position of the Zn 2p3/2 and Zn 2p1/2 peaks for the ZN/Ce–CN (Fig. 4(e)) are around 1022.2 eV and 1045.3 eV, respectively, demonstrating that the Zn element resides largely in the form of Zn2+ [12]. The Ce 3d spectrum is seen in Fig. 4f, which is associated with the Ce4+ 3d3/2 cerium state and has peaks with binding energies of 898.4 and 882.02 eV. The thick layer of g-C3N5 covering the ZN/Ce–CN surface is the reason for the faint Ce signal [40]. Hence, the ZnO/Ce–CN nanocomposite’s XPS spectrum revealed a robust chemical connection between ZnO and Ce–CN, which leads to the heterojunctions formation.

Fig. 4
figure 4

XPS spectra of ZN/Ce–CN photocatalyst. a Survey spectrum, (b) C 1s, (c) N 1s, (d) O 1s, (e) Zn 2p and (f) Ce 3d

Full size image

3.5 UV–Vis DRS

The optical characteristics of the as-obtained pure g-C3N5, ZnO, Ce–CN, ZN/CN, and ZN/Ce–CN nanocomposite were determined using UV–Vis DRS (Fig. 5a). ZnO’s absorption edge is ~397 nm, while g-C3N5’s is ~725 nm. g-C3N5 has a higher visible-light absorption capacity than pure ZnO, indicating that it can be activated by visible light and that it is a photo-responsive material capable of improving ZnO’s light absorption capability. Because of an extended π conjugated network resulting from the overlap between N 2p in a heptazine π conjugated system and N 2p orbitals of bridging azo moieties, g-C3N5 absorbs substantially in the visible region, with band tails reaching 660 nm [34]. While the absorption edge of Ce–CN is roughly 713 nm, compared to pure g-C3N5, the Ce–CN absorption band has a blue shift, which could be caused by the quantum confinement effect and Ce–CN’s defect state following cerium doping [41]. However, when compared to pure ZnO, the visible light absorption of ZN/Ce–CN nanocomposite increased considerably, and a large redshift was observed. It suggests that the nanocomposite are good photocatalytic materials. The Tauc model was utilized to further estimate the band gap values of the photocatalysts [42], as demonstrated by the Tauc plot presented in Fig. 5b. For g-C3N5, ZnO, Ce–CN, ZN/CN, and ZN/Ce–CN, the observed band gaps (Eg) were 1.96, 2.71, 1.77, 2.66, and 2.31 eV, in that order. It is apparent that combining ZnO with g-C3N5 only slightly broadens its light absorption spectrum. In contrast, coupling ZnO with Ce-doped g-C3N5 significantly enhances the light absorption capacity of the ZN/Ce–CN composite (with the band gap reduced to 2.31 eV), which supports higher photocatalytic activity.

Fig. 5
figure 5

UV–Vis diffuse reflectance spectra (a) and Tauc plot (b) of pure g-C3N5, ZnO, Ce–CN, ZN/CN and ZN/Ce–CN

Full size image

3.6 PL spectra analysis

Figure 6 depicts the optical photoluminescence (PL) emission spectra excited at 389 nm, which were utilized to determine the carrier transfer efficiency in the photocatalytic process. The noticeable fluorescence emission peaks of g-C3N5 indicate significant photogenerated electrons and holes recombination rates. The photoluminescence (PL) emission intensity of the ZN/CN composite is marginally reduced relative to that of pure ZnO. Nonetheless, the ZN/Ce–CN composite, which is formed by doping g-C3N5 with Ce and then combining it with ZnO, shows a significantly diminished PL emission peak. This pronounced reduction highlights the critical role of Ce doping in enhancing the transport and separation of photoinduced charge carriers within S-scheme heterojunction, effectively curbing carrier recombination [43].

Fig. 6
figure 6

photoluminescence spectra of pure g-C3N5, ZnO, Ce–CN, ZN/CN and ZN/Ce–CN

Full size image

3.7 Photo-electrochemical properties

The effectiveness of charge separation was further investigated using transient photocurrent response and electrochemical impedance spectroscopy (EIS). In general, the greater the intensity of the photocurrent, the greater the electron-hole separation and migration efficiency. The photocurrent transient response curve (I-t) (Fig. 7a) shows that when exposed to radiation, the photocurrent generated by ZN/Ce–CN are substantially stronger than that of Ce–CN and ZnO. The reason for this is that Ce–CN, in its role as an electron acceptor, is capable of efficiently capturing photoinduced electrons that are produced as a result of ZnO excitation, which in turn facilitates the separation and migration of photogenerated charge carriers [44].

Fig. 7
figure 7

a I-t curves and (b) EIS response of pure g-C3N5, ZnO, Ce–CN and ZN/Ce–CN

Full size image

Furthermore, as shown in EIS Fig. 7b. The radius of the arc is closely related to the transport efficiency of the photogenerated carriers [45]. The smaller arc radius of Ce–CN than g-C3N5 confirmed enhanced charge migration rate proficiency, signifying that Ce doping in g-C3N5 framework exhibited moderately poorer charge transferal resistance at the contact interface. Also, the incorporation of Ce into g-C3N5 significantly diminishes the recombination of h+/e− carriers. This is attributable to the fact that when Ce atoms are in a higher oxidation state (Ce(IV)), the vacant 5d orbitals can accept electrons, thereby creating defects that serve as trapping centers for photogenerated electrons, enabling their effective separation. Furthermore, ZN/Ce–CN possessed the highest electronic conductivity, which might be attributable to the formation of sufficiently tight interfacial connection between ZnO and CN. Photoelectrochemical investigations have conclusively demonstrated that the establishment of heterojunction between ZnO and Ce–CN significantly enhances interfacial electron transfer.

3.8 Photocatalytic degradation activity of MB

As shown in Fig. 8a, the produced nanoparticles’ photocatalytic activity was assessed using MB dye as a model pollutant. Initially, the dye solution was agitated with the photocatalyst in the dark for 30 min to achieve adsorption-desorption equilibrium. The pristine ZnO NPs and pure g-C3N5 nanosheets exhibited poor photodegradation efficiencies, reaching only 69 and 55%, respectively after 90 min of irradiation. It indicates that the recombination rate of photo-induced electron-hole pairs is faster in the one-component system of ZnO and g-C3N5. In comparison to single-component materials, all doped and composite photocatalysts demonstrated noticeably improved photocatalytic performance. ZN/(0.25%)Ce–CN nanocomposite shown the highest MB degradation efficiency of 97%, which was relatively higher than that of Ce–CN(76%), ZN/CN(77%), ZN/(1%)Ce–CN(92%), and ZN/(2%)Ce–CN(86%), under 90 min of UV-visible light irradiation. As a result, the effective production of heterojunction under the synergistic effect of ZnO and Ce–CN plays a critical role in MB degradation.

Fig. 8
figure 8

Photoreduction of MB over the as-prepared nanocomposite samples (MB = 20 mg L−1, catalyst = 0.75 g L−1): (a) time-dependent reduction curves; (b) pseudo first-order kinetic curves; (c) time-dependent reduction curves in different PH solutions

Full size image

Next, Fig. 8b shows that the catalyst’s degradation time of MB and -ln(C/C0) is proportionate, in line with the first-order kinetic equation [46]. The rate constant of (0.25%)ZN/Ce–CN nanocomposite was 0.0312 min−1, which is 5 times higher than that of g-C3N5, 2.9 times higher than that of ZnO, 3 times higher than that of Ce–CN, and 2.5 times higher than that of ZN/CN. The results revealed that (0.25%)ZN/Ce–CN nanocomposite had a high photocatalytic degradation efficiency for removing MB.

The role of pH on UV–Vis assisted degradation of MB in the presence of ZN/Ce–CN nanocomposite is shown in Fig. 8c. At pH 13, ZN/Ce–CN nanocomposite exhibited the most photocatalytic activity, reducing MB to around 99%. The ZN/Ce–CN nanocomposite degraded from 99 to 42% when the pH was lowered from 13 to 2, suggesting that the alkaline environment aids in the adsorption and photodegradation of MB. This is because the active sites of ZN/Ce–CN nanocomposite amass a significant amount of H+ ions in the pH range of 2–4.8, rendering them positively charged and repelling MB (a cationic organic molecule). However, with a pH of 10–13, there will be a lot of negative charge on the surface of the photocatalyst, which increases the strong electrostatic attraction between the ZN/Ce–CN nanocomposite and the positively charged MB and promotes photocatalytic efficiency [47].

3.9 Determination of reusability of composite ZN/Ce–CN

One of the most important features for degrading organic dyes is the material’s stability. To evaluate the stability of ZN/Ce–CN, the recycled photodegradation removal experiment of MB was studied. The photocatalytic degradation efficiency of ZN/Ce–CN composite stayed constant at 87.67% after five cycles (as seen in Fig. 9a), varying by about 2 to 4% per cycle, suggesting that it still possesses high photocatalytic activity. In addition, The XRD measurement results (Fig. 9b) demonstrated that the fresh ZN/Ce–CN and the utilized ZN/Ce–CN have similar XRD signals, indicating that ZN/Ce–CN had good structural stability. The results of this experiment revealed that the photodegradation efficiency of the ZN/CN nanocomposite slightly decreased after cycling experiments, which could be due to the fact that the photocatalyst was washed with distilled water several times and used for a long time, reducing their surface area and thus limiting the arrival of MB to the active site. However, it does not affect the good stability and reusability of ZN/CN in the photocatalytic degradation process.

Fig. 9
figure 9

a Repeatability cycles for MB degradation; b XRD spectra of ZN/Ce–CN before and after recycling experiments

Full size image

3.10 Trapping of reactive species

In order to identify the primary active species, free radical scavenger experiments were also used. Potassium iodide (KI), isopropanol (IPA), and p-benzoquinone (p-BQ) were respectively added to capture h+ (holes) [48], ·OH (hydroxyl radical) [49] and ·O2− (superoxide radical) [50].

According to Fig. 10a, without the use of a trapping agent, the degradation rate of MB was 97%, but it reduced dramatically with the addition of potassium iodide (KI), from 97 to 52%, showing that the h+ play a prominent role in the photocatalytic degradation process. The degradation rate of MB reduced from 97 to 67% when isopropanol (IPA) was introduced, demonstrating that ·OH constitute a major component impacting degradation. Additionally, the degradation rate of MB dropped marginally from 97 to 81% with the addition of p-benzoquinone (p-BQ) solution to the reaction solution, indicating that only a small portion of the degradation reaction is caused by ·O2−. According to the findings, the principal active ingredient is h+ in the photocatalytic degradation reaction, while ·OH also play a significant role, with some contribution from ·O2−.

Fig. 10
figure 10

a Active species trapping experiments over ZN/Ce–CN; ESR spectra in different time periods of TEMPO-h+ (b); DMPO-·OH (c); DMPO-·O2− (d)

Full size image

To confirm that the degradation process generates the aforementioned reactive oxidants, a possible free radical detection was performed utilizing an ESR test using DMPO as a probe (Fig. 10b–d). Under dark conditions, h+ showed a strong activity signal, which gradually decreased with increasing light time, due to the oxidation of TEMPO to TEMPO+ by h+, which has a strong oxidizing ability [51]. It can be seen that there are h+ in the degradation reaction, and the ZN/Ce–CN nanocomposite can maintain a significant quantity of h+ in the light. It implies that the h+ species plays the most important function in the photocatalytic process.

Figure 10c, d shows that under dark conditions, ZN/Ce–CN had almost no active species characteristic signals, but after 10 min of light irradiation, DMPO-·O2− and DMPO-·OH appeared as significant characteristic peaks, producing more photogenerated carriers, which provides significance for the photocatalytic reaction [52].

3.11 Band structure of ZN/Ce–CN composite and charge transfer mechanism

The Mott-Schottky technique was used to determine the semiconductor type and conduction band minimun (ECB). The Mott-Schottky curves (Fig. 11a, b) for ZnO and Ce–CN have positive slopes, suggesting that they are both n-type semiconductors [53]. In the meantime, the flat band potentials (Efb) of ZnO and Ce–CN were measured to be −0.49 V and −0.59 V versus Ag/AgCl electrode, corresponding to −0.29 V and −0.39 V versus normal hydrogen electrode (NHE), respectively. Furthermore, the gap between Efb and ECB for n-type semiconductors typically falls within the range of 0.1 to 0.3 eV and the ECB is more negative compared to the Efb [28, 54]. In this case, voltage difference between ECB and Efb is specified as 0.1 eV. That is, the ECB of ZnO and Ce–CN were −0.39 V and −0.49 V versus NHE, respectively. Additionally, the Eg values of ZnO and Ce–CN were 2.71 and 1.77 eV, respectively, according to Fig. 5b. As a result, their corresponding valence band maximum (EVB) were 2.32 and 1.28 V, respectively.

Fig. 11
figure 11

Mott-Schottky plots of ZnO (a) and Ce–CN (b)

Full size image

Together with the Eg values of Ce–CN and ZnO in the DRS diagram, the band potentials form Mott-Schottky test (Fig. 11) and the Efb of n-type materials often indexed their Fermi levels [55], a potential mechanism is conjectured (Fig. 12). For Ce–CN and ZnO, the CB potentials (versus NHE, pH = 7) are −0.49 and −0.39 V, respectively, and the VB potentials are 1.28 and 2.32 V, respectively. Ce–CN could form S-scheme or a type II heterojunction with ZnO because its CB and the VB potentials were more negative than ZnO’s [56]. Which was it, out of the two? A lot relied on the specific electron transfer mechanism in this case. The h+ produced on ZnO’s VB would flow to Ce–CN’s VB when exposed to light if a typical type II heterojunction formed (Fig. 12, left). As a result, the photoinduced holes were prone to gather on Ce–CN, while the VB of Ce–CN was found to be 1.28 V, which is both lower than the redox potential of H2O/·OH (2.27 V, versus NHE) and OH−/·OH (1.99 V, versus NHE) [29], indicating that the generation of ·OH impossibly occurred in the system. This finding contradicts the results of active species trapping experiments. Consequently, it is possible to rule out the creation of type II heterojunction in ZN/Ce–CN composite (Fig. 12, left).

Fig. 12
figure 12

S-scheme processes: photocatalytic degradation of MB in the presence of ZN/Ce–CN composite

Full size image

Additionally, in the S-scheme heterojunction (Fig. 12, right). Upon contact between Ce–CN and ZnO, electrons migrate from the higher Fermi level of the photocatalyst (Ce–CN) to the lower Fermi level (ZnO), thereby establishing an equilibrium between the Fermi levels at the interface of the two photocatalysts [28]. Once equilibrium is achieved, an accumulation layer and a depletion layer form at the interface between ZnO and Ce–CN, creating an internal electric field (IEF) that points from Ce–CN to ZnO [30]. The adjustment of Fermi levels (close to Efb) to a same position initiates the upward bend of Ce–CN and downward bend of ZnO, thus facilitating the complexation of electrons from ZnO and holes from Ce–CN under the IEF. Subsequently, the highly oxidizing holes gather on VB of ZnO (2.32 V) transform H2O or OH− into ·OH (H2O/·OH = 2.27 V, OH−/·OH = 1.99 V), while the highly reducing electrons accumulate on CB of Ce–CN (−0.49 V) capture O2, resulting in the formation of ·O2− (O2/·O2− = −0.33 V) [53]. Ultimately, the MB and its degradation intermediates were oxidized by the generated reactive species h+, ·OH, and ·O2−. The results of reactant scavenging tests and ESR spin trapping confirm the aforesaid observations. The above analyses all point to the creation of S-scheme heterojunction in the ZN/Ce–CN composite. ZN/Ce–CN composite can enhance the photocatalytic degradation efficiency of dyes by reducing the recombination rate of electron/hole pairs and improving the response to visible light. Firstly, Ce doping introduces defects into Ce–CN, which act as trapping centers for photo-induced electrons, capturing photogenerated electrons and facilitating the effective separation of holes and electrons. Secondly, the formation of a S-scheme heterojunction between Ce–CN and ZN further promotes the transport of holes and electrons. Finally, the synergistic effect of Ce doping and the S-scheme heterojunction promotes the generation of a large number of free electrons on the surface of ZN/Ce–CN, which can effectively catalyze the degradation of MB and significantly enhance the overall catalytic activity of the system.

4 Conclusion

The effective, ecologically acceptable, and stable photocatalytic material ZN/Ce–CN was successfully created using a simple calcination-hydrothermal technique for the photodegradation of MB. The ZN/Ce–CN heterojunction photocatalyst shown better photocatalytic activity, with a degradation rate of up to 97% at 90 min when compared to pure ZnO, g-C3N5, Ce–CN, and ZN/CN. The kinetic constant for MB degradation on the optimized ZN/(0.25%)Ce–CN heterojunction photocatalyst was 2.9 times higher than that of ZnO and 5.0 times higher than that that of g-C3N5. Cycling experiments demonstrated that the ZN/Ce–CN heterojunction photocatalyst retained good stability and repeatability after five cycles. ZN/Ce–CN has lower charge transfer resistance and higher charge separation efficiency because of the heterojunction formed between ZnO and Ce–CN. The incorporation of Ce significantly enhances the optoelectronic performance, playing an indispensable role. Active trapping experiments verified that h+ and ·OH play a substantial role for degradating MB. Finally, the Mott-Schott test further revealed that the presence of S-scheme heterojunction in the ZN/Ce–CN composite. This work demonstrates the significance of the synergistic effect engendered by cerium doping and the creation of heterojunctions between two distinct semiconductors, which is crucial in determining the photocatalytic properties.