3.1. Structure and morphology analysis
The typical XRD patterns of pure FeOCl, pure g-C3N4 and g-C3N4/FeOCl nanocomposites with different mass ratios are shown in Fig. 1. It is obvious that that pure g-C3N4 displays one characteristic peak at 2θ = 27.5°, which are associated with (002) diffraction planes of g-C3N4. For pure FeOCL, the diffraction peaks at peaks appear at 2θ = 11.16°, 26.10°,35.37°, 37.89°, 48.10°, and 61.10°, which exactly assigned to the crystal planes (010), (110), (021), (111), (200), and (221) of the hexagonal structure of FeOCl with JPCDS (No.72-0619), respectively. The characteristic peak shape of FeOCl is sharp and the intensity is high from the figure, indicating that the prepared FeOCl material has a complete crystal form and high crystallinity. In addition, the diffraction peaks of g-C3N4/FeOCl nanocomposites with different mass ratios are consistent with pure FeOCl, with the introduction of g-C3N4, no obvious g-C3N4 diffraction peak is observed, which may be attributed to the strong interaction between FeOCl and g-C3N4, resulting in the predominant (002) diffraction peak of g-C3N4 (2θ = 27.5°) coincides with the peak at 26.10° of FeOCl. Compared with pure FeOCl, no other impure peaks are observed for g-C3N4/FeOCl composites, indicating that the g-C3N4/FeOCl samples have been successfully synthesized.
The features of the as-synthesized photocatalysts were examined using SEM and HRTEM. From Fig. 3 (a, b), it is clear that the shape of pure FeOCl mainly exhibits the layered structure of cuboid nanorods, the average thickness of the nanorod structure is about from 300 to 400 nm. Fig. 3 (c), the HRTEM image of g-C3N4/FeOCl-2 calculates the characteristic d spacing as 0.275 nm and 0.236 nm are assigned to (120) and (111) plane of FeOCl. Fig. 3(d), the selected area electron diffraction of g-C3N4/FeOCl-20 shows that there are five obvious diffraction rings respectively corresponding to several representative crystal planes of FeOCl (110), (210), (200), (002), (152), indicating that the crystallinity of the composite sample has not decreased.
As shown in Fig. 4 (a), irregular black long strips are stacked together, covered with a thin layer of transparent yarn-like substance, indicating the successful synthesis of g-C3N4/FeOCl-2 nanocomposite. In Fig. 4 (b) shows the EDX spectra of g-C3N4/FeOCl-2 nanocomposites. It can be seen g-C3N4/FeOCl-2 nanocomposites only contain C, N ,O ,Cl, and Fe elements characteristic peaks. And the atomic percentage and weight percentage of related elements of the g-C3N4/FeOCl-2 nanocomposite are almost close to the nominal stoichiometry used in their respective precursors, without any impurity elements introduced. In addition, from Fig. 4 (c-h) TEM-mapping can be seen that the Fe, O, Cl, C, and N elements are distributed very uniformly, indicating that the g-C3N4/FeOCl-2 prepared by this method has high purity and compact structure.
3.2 XPS analysis
Fig. 5 (a) is the XPS full spectrum of g-C3N4/FeOCl-2 material. The peaks of Fe, O, Cl, C, and N elements can be clearly observed in the XPS full spectrum, indicating Fe, O, Cl, C, and N elements existed in the g-C3N4/FeOCl-2 material. In Fig. 5 (b), The binding energies of 2p3/2 and 2p1/2 of Fe are 708.7 and 722.4eV, respectively (Y. Q. Chen et al., 2018). Among them, the characteristic peak corresponding to the binding energy at 708.7 eV is Fe (II) 2p3/2 (Luo et al., 2019), which indicates that in addition to Fe(III) in the FeOCl sample, there is also a small amount of Fe(II). Fig. 5 (c) shows that the binding energies of O 1s are 527.5 and 530.2 eV, one peak at 527.5 eV is considered to be O attached to the metal (Fe-O), while the other at 530.2 eV is attributable to O adsorbed on the H2O molecule on the catalyst surface. Fig. 3 (d) shows that the binding energies of Cl 2p are 195.2 and 197.8 eV corresponding to the Fe–Cl bond of FeOCl. Fig. 5 (e) the binding energies of C1s are 286.1 and 281.3 eV, comes from sp2 N-C and sp2 C-C bond, respectively (Ma et al., 2017). In Fig. 5 (f), the peak at 398.2 eV was allocated to N element in tertiary N (N−C) (Asadzadeh-Khaneghah et al., 2018).
3.3. UV-vis DRS and FT-IR analysis
UV-vis DRS to test the visible light absorption characteristics and energy band structure of pure g-C3N4, FeOCl, and composite g-C3N4/FeOCl. As seen from Fig. 6(a), the visible light response range of bare g-C3N4 was 250−456 nm. Compared with pure g-C3N4, the light response range of pure FeOCl and composite g-C3N4/FeOCl both presented broad absorption over the whole region from 250 to 685 nm. Moreover, the absorption edges of g-C3N4/FeOCl-1, g-C3N4/FeOCl-2, and g-C3N4/FeOCl-3 nanocomposites are approximately located at 684, 685, and 676 nm in the visible light region, respectively. The visible light response range of the g-C3N4/FeOCl nanocomposite hasn’t changed than pure FeOCl, indicating that the introduction of g-C3N4 has little response to changing the visible light of the FeOCl. The bandgaps were also estimated using the model proposed by Tauc:\({(\alpha hv)^{1/n}}=A(hv - Eg)\) (Fig. 6(b)) (Huang et al., 2021; Varaprasad et al., 2021), the band gaps of g-C3N4 and FeOCl were 2.72 and 1.82 eV, respectively. And The bandgap of g-C3N4/FeOCl sample is basically the same as that of pure FeOCl. It shows that there is no obvious blueshift and redshift in the composite sample.
FT-IR analysis was used to characterize the information of chemical bonds and functional groups in the nanocomposites. As demonstrated in Fig. 7 broad band was observed at 3300 cm−1 in the spectrum belongs to the O─H of the H2O molecules that adsorbed on the surface of the material. At the position of 1200~1600 cm−1 of the peaks was from the C-N and C=N bonds in g-C3N4 (Ren et al., 2015). The Fe–O bond displays the absorption band at 811.2 cm−1 in the g-C3N4/FeOCl-2 system, indicating that there is not only the characteristic peak of g-C3N4 (Qu et al., 2019). Hence, the FT-IR measurement revealed the existence of g-C3N4, and FeOCl of functional groups in the g-C3N4/FeOCl-2 system, which is consistent with the results of XPS analysis.
3.4. Photoelectrochemical analysis
To investigate the photogenerated charge carrier transfer dynamics by measuring electrochemical impedance spectroscopy (EIS) and transient photocurrent spectra. As well known, the radius of the arc in the EIS plots corresponds to the separation efficiency of photogenerated electron-hole pairs, the arc radius is smaller that meant the better the separation effect of electron-hole pairs to lead to the better the photocatalytic performance. As exhibited in Fig. 8 (a), it can be clearly seen that the arc radius of g-C3N4/FeOCl-2 is the smallest, suggesting that the electron-hole pair separation effect of g-C3N4/FeOCl-2 is stronger than pure FeOCl and g-C3N4. Therefore, a higher high-speed charge carrier transfer at the g-C3N4/FeOCl-2 electrodes. The transient photocurrent test is to detect the number of photogenerated electron-hole pairs generated on the surface of the catalyst. The stronger the photocurrent intensity, the more photogenerated electrons are generated. In Fig. 8 (b) show the increased transient photocurrent of g-C3N4/FeOCl-2 compared with pure FeOCl, indicating that the compound with g-C3N4 can promote the separation of e− and h+ of FeOCl.
The type of semiconductor and flat-band potential can be measured by the Mott-Schottky test. From Fig. 9(a, b), it can be seen that the slopes of the Schottky curve of the FeOCL and g-C3N4 are positive slope, which impled that both FeOCL and g-C3N4 are n-type semiconductors. According to electrochemical measurements of calomel electrodes at 3000 Hz and 5000 Hz, the flat band potentials of FeOCl and g-C3N4 are +0.037 V and -1.023 V vs SCE, respectively. Then, the measured flat band potentials vs SCE were converted to the normal hydrogen electrode (NHE) scale by ENHE = ESCE + 0.059pH+0.242(pH=7)(Li et al., 2021). It was known that the flat band potential of an n-type semiconductor is approximately equal to CB. The CB position of FeOCl and g-C3N4 could be calculated to be +0.692 and -0.368 V. It can be calculated that the valence band (VB) energy levels of FeOCl and g-C3N4 were +2.512 and +2.352 eV by the empirical formula ECB = EVB – Eg (Zheng & Zhang, 2018).
3.5 Photocatalytic activity and Radical-trapping experiment
The photo-Fenton catalytic activity of g-C3N4, FeOCl, and g-C3N4/FeOCl samples was verified by degrading RhB solution in 90min (Fig. 10). All of the experiments were conducted by adding 0.4 mL H2O2 and the scale of pH=6~7. According to the results, FeOCl and g-C3N4 have lower RhB removal under dark conditions compared with composite samples, indicating relatively poor adsorption efficiency of RhB on the as-prepared pure FeOCl and g-C3N4. Under 90 minutes of light, the degradation rates of g-C3N4/FeOCl-1, g-C3N4/FeOCl-2, g-C3N4/FeOCl-3 nanocomposite are 80.7%,92.2% and 88.0% respectively. However, the removal rates of pure FeOCl and g-C3N4 are 86.7% and 45.1%. The ability of the composite material to degrade pollutants is significantly improved. g-C3N4 is attached to the surface of the long FeOCl, and the full contact between the two forms a Z-type heterojunction (Zhao et al., 2020). Promote the migration and separation of photo-generated carriers.
In order to explore the photo-Fenton active materials of the g-C3N4/FeOCl-2/vis/H2O2 system, which was added active materials capture agents such as isopropanol (IPA), methanol (IA) and benzoquinone (BQ) by radical-trapping experiment (Fig. 11(a)). The degradation rate of RhB was significantly reduced after adding IPA in the suspension. And h+ also inhibited the progress of the experiment because h+ could capture enough H2O to produce ·OH. The degradation rate of RhB didn’t significantly change after adding BQ in the suspension, showing that the photo-induced ·O2− little influenced the degradation of RhB from Fig. 11(b). The radical-trapping experiment results demonstrate that ·OH plays a most important role, h+ is secondary species, and ·O2− also participated in the degradation of RhB.
3.6 Photocatalytic mechanism
Combining the above results, both g-C3N4 and FeOCl were n-type semiconductors and the CB and VB energy levels of FeOCl are higher than g-C3N4. Hence, the recombination of g-C3N4 and FeOCl would form a Z-type heterojunction. When the g-C3N4 and FeOCl contacted, electrons would spontaneously diffuse from g-C3N4 with a high Fermi level to the FeOCl from in Fig. 12 (a) (Yang, 2021). At the same time, a built-in electric field will be formed at the interface and the CB and VB of g-C3N4 and FeOCl would move which was forming a side of heterojunction, eventually reaching a state of thermal equilibrium. The photo-Fenton mechanism diagram of g-C3N4/FeOCl could be drawn that when sunlight irradiates the surface of the catalyst, the CB and VB of g-C3N4 and FeOCl would produce non-equilibrium carrier electron-hole pairs. The heterojunction of g-C3N4/FeOCl and the state of thermal equilibrium were broken. The CB of FeOCl electrons migrated to the VB of g-C3N4 to recombine with holes, which the electrons were gathered on the CB of g-C3N4 and a large number of holes produced on the VB of FeOCl.
According to Fig. 12 (b), H2O2 was converted ·OH and OH− by the electrons that gathered at the CB of g-C3N4, the VB of FeOCl h+ captured H2O to convert ·OH. During this process, the H2O2 and H2O were served as the electron and hole acceptors which the recombination of holes and electrons further were successfully limited (Xing et al., 2020). On the other hand, on the surface Fe3+ of FeOCl material transformed into the Fe2+ with the existence of H2O2 under the radiation of sunlight condition, and the Fe2+ was easily reacted with H2O2 to generated the ·OH (Ye et al., 2018). In addition, owing to Standard reduction potential (E0(O2/·O2−)) was more positive than the CB of g-C3N4 and Standard oxidation potential (E0(·OH/OH−)) was more negative than the VB of FeOCl, the electron on the CB of the g-C3N4 would also reduce O2 to form the ·O2−and the h+ which produced on the VB of the FeOCl could oxide the ·OH− into the ·OH (Chen et al., 2020). Based on the radical-trapping experiment obtained the possible photo-Fenton degradation mechanism.