3.1. Structure and morphology
The structure and crystalline phase were analyzed by XRD. Fig. 1a confirmed g-C3N4 has two diffraction peaks at vicinity of 12.9° (100) and 27.5° (002), which were in good agreement with the standard XRD pattern of graphitic carbon nitride (JCPDS 87-1526). The peak at 12.9° was confirmed by the in-planar tri-s-triazine structural ordering of the conjugated aromatic system, and the peak at 27.5° was represented the inter-planar periodic lamellar ordering of typical graphite-like carbon nitride (Li et al. 2016). According to Bragg’s Law, the distance between the in-plane layers was about 0.69 nm, and the distance between the inter-planar layers was about 0.33 nm (Lu et al. 2015; Zou et al. 2019). On the contrary, the intensities of both two peaks were significantly abated, and the peaks were minor shifted in 10-Cd-g-C3N4 samples, indicating that the Cd elements incorporated into the g-C3N4 lattice and resulting to the reduced hydrogen bond effect (Sher et al. 2021), as a consequence, the distance between the inter-planar layers of g-C3N4 was increased (Thaweesak et al. 2017; Yu et al. 2016). These findings confirmed the successful doping of Cd, and the introduction of Cd decreased the crystallinity of g-C3N4 sample (Abu Hanif et al. 2021). Apparently, there were no peak of cadmium in 10-Cd-g-C3N4 sample, which was due to the small content of cadmium and high dispersion in solution (Jia et al. 2019; Wang et al. 2020b).
The functional groups in the g-C3N4 and 10-Cd-g-C3N4 were determined by FT-IR. As illustrated in Fig. 1b, the pure g-C3N4 presented numerous absorption bands. The peak at 809 cm−1 was assigned to the typical out-of-plane bending vibration of tri-s-triazine-based structure (Liu et al. 2020), which verified the presence of triazine units (Wang et al. 2018a; Yan et al. 2009b). Analogously, five characteristic peaks were found in the region of 1200–1700 cm−1, which appeared at 1245, 1327, 1412, 1573 and 1645 cm−1, respectively, indicating the distinctive stretching mode of aromatic C–N heterocycle (C6N7) (Park et al. 2011; Sobhana S S et al. 2011). Similarly, the wide region from 3000 to 3500 cm−1 were ascribed to the stretching mode of the partial hydrogenation of exposed nitrogen (N-H) and the adsorbed water molecules (O-H) (Tonda et al. 2014; Wang et al. 2018a; Yan et al. 2016). Obviously, the 10-Cd-g-C3N4 sample exhibited almost similar FT-IR features with g-C3N4, indicating the primary framework of g-C3N4 was well preserved. Specifically, a weak peak was found in the region of 2150 cm−1, which may assign to the additional stretching mode caused by the interaction of cadmium and g-C3N4 (Sher et al. 2021). Additionally, the stretching mode of the partial hydrogenation of exposed nitrogen (N-H) was shifted and weakened, which corresponded to the Cd dopant was incorporated into g-C3N4 (Yang et al. 2013; Li et al. 2016).
The morphologies of as-prepared samples were examined by SEM. Fig. 2a and Fig. 2b presented the image of pure g-C3N4, which were layers and porous structure piled up in an irregular manner (Narkbuakaew & Sujaridworakun 2020). Also, there were large cross-sectional area between layers, which were excellent doping sites (Sher et al. 2021). After Cd doping, 10-Cd-g-C3N4 sample was composed of thin layer structure stacking with scattered crystals particles Fig. 2(c-d), revealing the aggregated nanosheets of g-C3N4 was exfoliated, the crystallinity of g-C3N4 was decreased (Abu Hanif et al. 2021). Elemental mapping illustrated that Cd element was homo-dispersed at the surface of g-C3N4, suggested the success doping of Cd (Figure S1, supporting information). The elements weight ratio of 10-Cd-g-C3N4 sample were illustrated in Table S2, where the elements C was 25.73%, N was 61.98% and Cd was 12.28%, indicating the low concentration of Cd compared with others elements, which confirmed the cadmium doped on the g-C3N4 with a slight atomic ratio. The energy dispersive X-ray spectra (EDS) were illustrated in Fig. S1a, and only three peaks (C, N, Cd) were presented in the EDS mapping spectrum, confirming that the C, N, Cd elements were homogeneous distribution and the prepared 10-Cd-g-C3N4 sample was highly pure.
The detailed morphologies and microstructure were further surveyed by TEM. All the prepared samples were nanometer range particle size. As depicted in Fig. 3a and Fig. 3b, it could be apparently seen that the pure g-C3N4 was transparent or non-transparent, owing to the overlap of multilayered nanosheets. Simultaneously, as confirmed in Fig. 3c and Fig. 3d, the 10-Cd-g-C3N4 exhibited an ultrathin nanosheets structure, which was beneficial for increasing the specific surface area. Analogously, a typical silk nanosheet with smooth external surface of g-C3N4 was observed, which manifested that the graphene-like layers were uniformly distributed and randomly oriented. Obviously, the elemental mapping (Fig. S1 and Fig. 3(e-i)) confirmed the cadmium elements were uniform interspersing at the surface of g-C3N4, and exhibited high intercontact with g-C3N4 due to the electrostatic attraction. The results were consistent with previous study that the electron-rich Cd or CdS clusters were firmly anchored on the electron-unsaturated g-C3N4 surface, which illustrated that the cadmium was doped on the g-C3N4 surface (Zhang et al. 2016). Fig. 3i illustrated the amorphous structure of 10-Cd-g-C3N4 sample. The selected diffraction pattern (Fig. 3j) shows two distinct amorphous diffraction rings, which correspond to two peaks of carbon nitride in the XRD pattern and correspond to the XRD data. All these results clarified that 10-Cd-g-C3N4 samples were successfully constructed.
The BET specific surface area, pore volume and pore size distribution were tested by N2 adsorption-desorption and the results were depicted in Table S3 and Fig. S2. The specific surface area of 10-Cd-g-C3N4 was 16.46 m2/g, and the pore volume was 0.123 cm3/g, whereas the surface area of pure g-C3N4 was 13.467 m2/g, and the pore volume was 0.077 cm3/g. Obviously, the 10-Cd-g-C3N4 sample obtained a larger surface area and pore volume than pure g-C3N4. Analogously, the porosity of 10-Cd-g-C3N4 was obviously increased, which related to the cracks of light layer during crystal growth process or the evaporation of chlorine gas during heating process. Undoubtedly, larger surface area and pore volume could provide more reaction active sites, increase abundant light absorption area and accelerate the transfer of photogenerated carriers, accordingly, achieving a higher photocatalytic degradation rate (Chang et al. 2015). The N2 adsorption-desorption isotherm (Fig. S2a) of 10-Cd-g-C3N4 was exhibited a type IV curve with H3 hysteresis hoop, which represented the presence of slit shaped mesoporous structure (Iqbal et al. 2017).
The element chemical bonding states and electronic structure details were investigated by XPS analysis. Fig. 4a characterized the full survey spectrum of g-C3N4 and 10-Cd-g-C3N4. Obviously, the pure g-C3N4 sample was consisted of carbon (C), nitrogen (N), oxygen (O) element and the 10-Cd-g-C3N4 sample was consisted of C, N, O and cadmium (Cd) element. The presence of O element was caused by carbon nitride thermal polymerization process or catalyst surface absorption of water (Chou et al. 2016; Fang et al. 2016; Xue et al. 2019). Equally, no signal of chlorine (Cl) element (around 200 eV) could be observed, since Cl atoms were evaporated during heating process (Amirthaganesan et al. 2010).
The peaks of Cd 3d were situated at 405.6 eV and 412.4 eV in Fig. 4b, which were ascribed to the Cd 3d5/2 and Cd 3d3/2 (Reddy et al. 2021), indicating the existence of Cd2+ in 10-Cd-g-C3N4 (Zhang et al. 2016a), and further confirming the presence of Cd dopant (Abu Hanif et al. 2021). Similarly, Fig. 4c elucidated the high-resolution C1s spectra, two sharp peaks were situated at 284.8 and 288.3 eV, respectively. The former peak at 284.8 eV was specifically on behalf of carbon atoms (C–C bonds), namely graphite or amorphous carbon (Liu et al. 2010; Takanabe et al. 2010), and the latter sharper peak at 288.4 eV resulted from sp2 C atoms bonded with adjacent N atoms inside the aromatic structure (N–C=N) (Thomas et al. 2008). Obviously, three peaks could be perceived from the high-resolution N 1s spectra (Fig. 4d), the peak at 398.8 eV originated from sp2-bonded N atoms in triazine units (C=N–C) (Li et al. 2009b; Takanabe et al. 2010), the peak at 401.1 eV corresponded to amino groups (C–N–H) from the surface uncondensed bridging N atom (Liu et al. 2010; Thomas et al. 2008), which were similar to previously published literature (Chao et al. 2014; Matanović et al. 2015). The peak at 404 eV was assigned to the effects of surface charge localization in the heterocycles or the π–π* excitations between the stacking interlayers (Kong et al. 2018; Li et al. 2018; Zhang et al. 2014; Zhang et al. 2012). Both the high-resolution C 1s and N 1s spectra confirmed the presence of g-C3N4. Simultaneously, both C 1s and N 1s spectra of 10-Cd-g-C3N4 sample exhibited an upwards shifted compared with pure g-C3N4, all these positive shifts could be ascribed to the electrons transfer of 10-Cd-g-C3N4 sample, further proving the strong interaction between Cd elements and g-C3N4 (Ji et al. 2019; Yan et al. 2019).
3.2. Optical properties
The optical absorption characteristics was an evaluation criterion to optical properties and electronic band structures of catalyst. Simultaneously, the optical absorption range and capacity of catalysts affected the photocatalytic performance. Fig. 5a revealed the UV–vis diffuse adsorption spectrum of g-C3N4 and 10-Cd-g-C3N4. An absorption threshold at 465 nm was found in pure g-C3N4. Obviously, the absorption threshold of 10-Cd-g-C3N4 had a significantly redshift toward longer wavelengths compared with pure g-C3N4, which illustrated that Cd doping may form a new level in g-C3N4 and cause a narrower band gap (Abu Hanif et al. 2021). Moreover, the intensity of light adsorption was significantly enhanced in 10-Cd-g-C3N4 sample, which further confirmed the improvement of light absorption performance. Additionally, 10-Cd-g-C3N4 extended the adsorption edges to near 600 nm, this extension is conducive to increasing the light absorption of g-C3N4 and increasing the generation of electron–hole pairs (Chen et al. 2021). Simultaneously, the UV–vis diffuse adsorption spectrum of other samples with different cadmium contents was clarified in Fig. S3a. Among these samples, the 10-Cd-g-C3N4 sample displayed the highest visible light absorption in a long range.
Both g-C3N4 and 10-Cd-g-C3N4 sample are direct-gap materials. The band gap (Eg) of g-C3N4 and 10-Cd-g-C3N4 were calculated based on the related Tauc plots in Fig. 5b (Luo et al. 2017). Noticeably, the band gap of pure g-C3N4 was assigned to 2.57 eV (Kong et al. 2018; Wang et al. 2009b; Zhang et al. 2012), and the band gap of 10-Cd-g-C3N4 was 2.29 eV. Undoubtedly, the narrower band gap reduced the excitation energy for photogenerated carriers, leading to a significant improvement in the visible light response. As a consequence, Cd doping can significantly affect the light absorption performance of g-C3N4 (Ge et al. 2012).
The charge separation, migration and recombination were analyzed. Fig. 6a indicated the PL spectra of 10-Cd-g-C3N4 and g-C3N4 samples. Commonly, the weakened PL intensity meant enhanced photoinduced charge separation and transfer efficiency. Apparently, the strong peak at 460 nm in pure g-C3N4 was consistent with the absorption edge of the UV-vis DRS, and indicating the severe recombination of the photogenerated electron−hole pair (Yu et al. 2013). Equally, the highest intensity peak demonstrated a red-shifted from 460 nm to 490 nm in 10-Cd-g-C3N4, which was associated with the band gap narrowing effect (Gu et al. 2018; Zou et al. 2019). Analogously, the PL spectra of other samples with different cadmium contents was manifested in Fig. S3b. Obviously, the PL spectra of Cd-g-C3N4 sample significantly fell with the doping of cadmium, which could ascribe to the fast interfacial charge migration. Additionally, the PL spectra of 10-Cd-g-C3N4 was lowest, which illustrated that Cd doping restrained the recombination of photoinduced charge due to the strong contact between cadmium and g-C3N4 (Thomas et al. 2008).
Figure 6b was a Mott-Schottky plots with flat band potentials at 1000 Hz. Commonly, conduction band potential is equivalent to flat band potential for an n-type semiconductor. Both g-C3N4 and 10-Cd-g-C3N4 displayed a positive slope, indicating g-C3N4 and 10-Cd-g-C3N4 are both n-type semiconductors. Meanwhile, the Mott-Schottky slope of 10-Cd-g-C3N4 was abated compare with pure g-C3N4, owing to the increased electrical conductivity and the quick mobility of charge carriers, which revealing the electron donor density was increased in 10-Cd-g-C3N4 (Jun et al. 2013; Pan et al. 2018; Yang et al. 2013; Yuan et al. 2018; Zhou et al. 2014). As a consequence, higher electron donor density is helpful for improving photocatalytic performance (Zhou et al. 2014). The conduction band potential of g-C3N4 and 10-Cd-g-C3N4 concluded from the Mott-Schottky plots was -0.84 eV and -1.02 eV, compared with Ag/AgCl electrodes, corresponding to -0.64 eV and -0.82 eV vs normal hydrogen electrode (NHE).
Metal elements doping influenced the band edge of catalyst to a great extent, and the redox capacity of semiconductors were evaluated via the band edge position of conduction (ECB) and the valence (EVB) (Xiong et al. 2016). The EVB could obtained from following equation (Zhang et al. 2010):
E VB = ECB + Eg
Figure S4 revealed the EVB and ECB positions of as-prepared samples. For pure g-C3N4, the EVB was 1.93 eV, on the basis of the experimental Eg (2.57 eV), the ECB was -0.64 eV. Meanwhile, the EVB of 10-Cd-g-C3N4 was 1.47 eV and the ECB was -0.82 eV, according to the experimental Eg (2.29 eV).
Other photoelectrochemical techniques such as electrochemical impedance spectroscopy (EIS) and photocurrent response measurement were employed to investigate the movement of photoinduced electrons. Fig. 6c revealed the photocurrent responses of as-prepared samples, both g-C3N4 and 10-Cd-g-C3N4 testified an outstanding reproduceible photostability under successive on/off irradiation cycles. Homoplastically, the photocurrent response density of 10-Cd-g-C3N4 was significantly increased compare with pure g-C3N4, indicating the introduction of Cd element increased the conductivity of g-C3N4 and accelerated the separation of electron-hole pair (Ren et al. 2017). EIS was another method to evaluated the separation and transfer efficiency of photoinduced carriers. Commonly, the smaller arc radius suggested smaller transfer resistance, namely the better separation and transfer efficiency of photoinduced carriers (Lu et al. 2017; Yang et al. 2002). As depicted in Fig. 6d, the arc radius of 10-Cd-g-C3N4 was smaller than pure g-C3N4, which reflected that 10-Cd-g-C3N4 possessed high-efficiency photoinduced carriers separation ability and faster interfacial charge transfer level (Zhu et al. 2015).
3.3. Photocatalytic activities
To testing the effect of different content of Cd doping on degradation performance, a group of comparative degradation experiments were carried out under the simulated visible-light irradiation (λ > 420 nm). In this paper, TC was selected as contaminant because of the high stability under visible light. As illustrated in Fig. 7a, the degradation rate of TC without catalysts can be neglected. Apparently, the adsorption capacity of all Cd doped g-C3N4 samples were slightly larger than that of pure g-C3N4 during the dark stage, which corresponding to an increase in the specific surface area of the catalyst. Obviously, all Cd doped g-C3N4 samples exhibits great increased photocatalytic degradation performance after 60 min visible-light irradiation compared with pure g-C3N4 (43.9%), confirming the doping of Cd was truly enhanced the photocatalytic efficiency (Abu Hanif et al. 2021). Apparently, the 10-Cd-g-C3N4 sample demonstrated the significantly improved photocatalytic degradation performance (98.1%), which can be attributed to the effective separation and transfer of photogenerated charges originating from the photocatalyst interface. Additionally, the photocatalytic degradation performance within 10 min of 10-Cd-g-C3N4 sample was tested (Fig. S5a), the result illustrated that the TC was degraded almost 80% within 10 minutes. Table S4 illustrated the comparison with other similar types researches, we found the removal rate of this work was excellent, by compare this work with other similar types of work.
Furthermore, the degradation performance of Cd doped g-C3N4 samples first enhanced with the increase of Cd content. As the content of Cd increased, the utility efficiency of visible light was promoted, as a consequence, more photoinduced carriers were provided. Unfortunately, when the doping content of Cd continuously increased, the degradation performance was decreased, the photocatalytic degradation performance of 15-Cd-g-C3N4 sample was only 75.9%, indicating the excessive Cd element could serve as photo-carrier’s recombination center or blocked the surface active site to restrained the light absorption capacity and thus inhibiting photocatalytic degradation performance (Li et al. 2009a; Zhang et al. 2016b; Zhang et al. 2010).
The pseudo-first-order kinetic model could further investigate the degradation performance of the as-prepared catalyst, and the formula was expressed as follow:
ln (Ct/C0) = – Kappt
C0 is the initial concentrations, Ct is the concentrations at time t, and Kapp is reaction rate constant (min−1), respectively. Fig. S5b presented the pseudo-first-order kinetic plots of Cd-g-C3N4 samples and pure g-C3N4. Under the same condition, the Kapp of Cd-g-C3N4 samples were higher than pure g-C3N4 and the Kapp of 10-Cd-g-C3N4 sample was the highest, demonstrating that 10-Cd-g-C3N4 had the best photocatalytic performance, which was consisted with the above experimental results.
The effects of different reaction conditions, such as initial concentration of TC, catalyst dosage and pH, were studied to meet practical application. Fig. 7b displayed the influence of different catalyst dosages on TC photodegradation. Apparently, as the catalyst dosages increased between 0.6 and 0.8 g/L, the removal rate of TC was enhanced monotonically, which was ascribed to the increased contact active sites between contaminants and materials, and more contaminants were absorbed on the surface of materials. Apparently, the catalyst dosages at 0.8 g/L elucidated the best photocatalytic degradation performance. Nevertheless, when the catalyst dosages over 0.8 g/L, the removal rate of TC was abated slightly, the photocatalytic degradation efficiency was 82.3% when the catalyst dosages at 1.0 g/L, which owing to the excessive catalyst reduced the photo-adsorption ability, as a result, hindering the effective migration of charge carriers (Zhang et al. 2016a).
The effect of different pH on TC photodegradation was clarified in Fig. 7c, and diluted hydrochloric acid and sodium hydroxide were used to adjust the initial pH. Obviously, the best degradation efficiency of TC was 94.4% at pH=5. Consequently, it was concluded that degradation efficiency was decreased in low pH values and alkaline environment, which can be explained by the surface charge of contaminants and materials. Additionally, the effect of initial concentration of TC was illuminated in Fig. 7d. It could be notably detected that the removal rates were abated with the initial concentration increased, and the best removal rate was 96.7% at 10 mg/L of TC. Two possible reasons could be proposed to elucidate this tendency. Firstly, high concentrations of contaminants might accumulate at the surface of catalyst, inhibiting the light absorption capacity. Secondly, intermediate products produced in the process of contaminants degradation might occupy the surface active site, so that contaminants had no contact with the catalyst.
In order to evaluated the practicability for catalysts, the stability tests were carried out by reusing the catalyst. Fig. 8a depicted the degradation rates of TC in four consecutive cycles. Obviously, the degradation rates were slightly decreased with the reuse of 10-Cd-g-C3N4, which could be attributed to the catalyst loss during the collection process. Whereas, the degradation rates still maintained a high level (89.5%) at the fourth cycle. Simultaneously, Fig. 8b demonstrated the XRD analysis of catalyst, no obvious peak changes were found in used catalyst, indicating the outstanding stability of 10-Cd-g-C3N4.
The ability of mineralization of note is an important index in photocatalytic process, which is used to ascertain the degree of contaminants removal. As intriguingly indicated in Fig. S4b, the mineralization efficiency of TC reached to 34% within 60 min visible light irradiation, which confirmed that 10-Cd-g-C3N4 could actually degrade TC into small molecule intermediate compounds (CO2, H2O).
The trapping experiments were used to determine the active species participating in the degradation process (Li et al. 2015). EDTA-2Na, IPA and TEMPO were added as scavengers to capture h+, •OH and •O2−, respectively. Fig. 9a and Fig. 9b illustrated that the degradation rate of TC was suppressed by EDTA-2Na, indicating that h+ was the main active species of 10-Cd-g-C3N4. Similarly, the degradation rate of TC abated slightly when the TEMPO and IPA were added, revealing that •O2− and •OH played auxiliary roles in photocatalytic degradation process (Yan et al. 2010; Yu et al. 2018).
The ESR was carried out to verify the above results. As illustrated in Fig. 9c and Fig. 9d, it is hard to see the ups and downs signals of active species in 10-Cd-g-C3N4 under dark conditions. Paradoxically, the strong intensity 1:2:2:1 signals of •OH and 1:1:1:1 signals of •O2− radicals could be found under the visible light irradiation, which proving the visible light is prerequisite for 10-Cd-g-C3N4 to product •OH and •O2− active species.
According to all aforementioned results, the introduction of Cd element speeding up the separation and transfer of electron-hole pairs, decreasing the recombination rate of photogenerated charge carriers, providing more reactive sites and accelerating cross-plane diffusion in g-C3N4 nanosheets (Iqbal et al. 2017), thus achieving the enhanced photocatalytic property. A brief mechanism of 10-Cd-g-C3N4 sample during TC degradation was summarized in Scheme 1. The electrons and holes were firstly generated and separated under light irradiation, and the electrons transferred from valence band to conduction band spontaneously, as a consequence, leaving holes at valence band and electrons accumulated at conduction band (Eq. (1)). Then, the absorbed O2 reacted with electrons to formed superoxide radical (Eq. (2)). Similarly, the surplus electrons reacted with superoxide radical to formed hydrogen peroxide (H2O2) (Eq. (3)). As listed in Fig. S4a, the ECB of 10-Cd-g-C3N4 was negative than O2/•O2− (-0.33 eV vs NHE), as a result, the accumulated electrons could react with O2 to formed •O2−. But the EVB of 10-Cd-g-C3N4 was negative than OH−/•OH (2.40 eV vs NHE), consequently, the accumulated holes could not react with OH− to formed •OH, which was not consistent with ESR result. Consequently, the •OH probably derived from H2O2 decomposition (Eq. (4)). From what has been discussed above, •O2−, •OH and h+ were worked together to degrade TC into small molecules (Eq. (5)). The reaction equations involved in the photocatalytic reaction process are manifested below (Jia et al. 2020).
10-Cd-g-C3N4 + hv → e− + h+ (1)
O2 + e− → •O2− (2)
•O2− + e− + 2H+ → H2O2 (3)
H2O2 + e− → •OH + OH− (4)
•O2− + h+ + •OH + TC → H2O + CO2 + intermediate products (5)