Structures and morphologies of samples
Figure 1 displays the XRD patterns of the as-prepared samples. In Fig. 1a, the peaks at 2θ of 8.84°,18.4°༌27.5°, and 60.7° correspond to (002), (004), (006), and (110) crystal planes of Ti3C2 (Xu et al. 2018; Cai et al. 2018), respectively. In Fig. 1b, the diffraction peaks of the Ti3C2 and g-C3N4NS were presented (Xu et al. 2019). As shown in Fig. 1c–f, the CdS with cubic phase (JCPDS No 80 − 0019) is obtained in addition to Ti3C2 and g-C3N4NS phases.
Fig. 2 shows the Raman spectra of the samples. In Fig.2a, the Raman peaks located at 325.5 cm−1 (A1g mode) corresponds to the out-of-plane stretching vibrations of Ti and C (Hu et al. 2015), and the peaks at 1456 and 1519 cm-1 can be attributed to the D-band and G-band of Ti3C2(Naguib et al. 2014). In Fig.2b, the typical peaks at 301.1 and 600.2 cm-1 are assigned to the first-order LO Raman peak and second-order LO phonon vibrational mode of cubic CdS (Rengaraj et al. 2011), respectively. In Fig.2 c, the peak at 474cm-1 is referred as the Raman signal of g-C3N4NS(Sutar et al. 2020). As shown in Fig.2d, the Raman peaks of Ti3C2, g-C3N4NS and CdS are shown, respectively. The XRD and Raman results verify that the ternary composites are composed of Ti3C2, g-C3N4NS and CdS.
FigureS1 shows the morphologies of g-C3N4 and g-C3N4NS. g-C3N4 shows an irregularly wrinkled morphology, and the prepared g-C3N4NS presents a thin sheet shape. Figure 3 presents the morphologies of the Ti3C2, Ti3C2/g-C3N4NS and CdS/Ti3C2/g-C3N4NS samples. The prepared Ti3C2 exhibits a 2D-layered morphology (Fig. 3a). In Fig. 3b, the g-C3N4NS is tightly coated around the Ti3C2, forming a rough surface. A rough surface facilitates the transmission of photogenerated carriers (Zhang et al. 2020). As shown in Fig. 3c-f, the SEM images of CdS/Ti3C2/g-C3N4NS composites show that the CdS particles are deposited and tightly attached on the interlayer and surface of Ti3C2/g-C3N4NS. With the further increase in CdS amount, the interlayer of the composite occupies a large amount of CdS particles, and the 2D-layered structure of the ternary composite is weakned (Fig. 3f). The EDX spectra and the corresponding elemental mapping were carried out for the CTC-4:1sample (Fig.S2). Ti, C, N, S, and Cd elements are uniform distributed in the composite.
The HRTEM image of CTC-4:1 is shown in Fig. 4. The lattice spacings of 0.250nm and 0.336nm correspond to the (006) and (111) crystal facets in Ti3C2 and CdS(Cao et al. 2013; Ding et al. 2019), respectively. The g-C3N4NS with low crystallinity do not show obvious lattice fringes (Wang et al. 2019). A compact and continuous interface exists among Ti3C2, CdS, and g-C3N4NS, which can provide efficient separation and diffusion of photoinduced carriers (Kang et al. 2017).
Xps Analysis Of Samples
Figure 4 shows the XPS spectra of the CTN-4:1 and CdS. The XPS survey spectrum indicates that the CTN-4:1 is composed of Ti, C, N, S, and Cd elements (Fig. S3).
In the Ti2p XPS spectrum of CTN-4:1 (Fig. 5A), the peaks at around 454.7, 458.7 and 463.9 eV are attributed to the Ti-C, Ti2p3/2 and Ti2p1/2 in Ti3C2(Yang et al. 2018; He et al. 2020). For the C1s spectrum (Fig. 5B), the peaks at approximately 286.26 eV can correspond to C–Ti on the CTN-4:1 surface(Peng et al. 2017), while the peaks at 284.6 and 288.36 eV are attributed to the C-C and C-N = C bonds in g-C3N4NS (Diao et al. 2020).
The S2p and Cd3d XPS spectra of CTN-4:1 and CdS were compared. In the S2p and Cd3d spectra of CTN-4:1, peaks at 161.9 and 163.2 eV can be correspond to S2p1/2 and S2p3/2 of S2+ (Ai et al. 2019) (Fig. 5C), and the peaks at 412.3 and 405.5 eV can be ascribed to Cd3d3/2 and Cd3d5/2 of Cd2+(Ai et al. 2019) (Fig. 5D). Compared with CdS, the binding energies of S2p and Cd3d in the CTN-4:1 show a shift to the right, suggesting the strong interactions between CdS and Ti3C2/g-C3N4NS, forming a heterojunction composite(Wang et al. 2018; Wang et al. 2019).
Bet Analysis Of Samples
The Nitrogen adsorption-desorption isotherms of as-prepared samples are shown in Fig. S4. The specific surface area and pore volume values of all samples are summarized in Table 1. The surface area and pore volume values of obtained g-C3N4NS are evidently higher than that of g-C3N4. In contrast to the Ti3C2/g-C3N4NS, the CdS-loaded ternary composites increase the surface area and pore volume values. Among the ternary composites, the CTN-4:1 sample presents the largest surface area, which may provide more reactive active sites for photocatalytic degradation (Elfiad et al. 2018).
Table 1
Surface area and pore volume of the samples
Sample
|
Surface area (m2/g)
|
Pore volume (cm3/g)
|
g-C3N4
g-C3N4NS
CdS
Ti3C2
Ti3C2/g-C3N4NS
CTN-2:1
CTN-3:1
CTN-4:1
CTN-5:1
|
45.09
70.72
54.83
6.501
21.86
46.28
51.29
52.99
48.89
|
0.22
0.53
0.30
0.04
0.18
0.22
0.24
0.26
0.25
|
Photocatalytic Activity Of As-prepared Samples
Photocatalytic degradation of orange II
In Fig. S5, the adsorption and degradation curves of orange II exhibit that g-C3N4NS has higher adsorption and degradation performances than that of g-C3N4, which agrees well with the BET analysis. Figure 6A presents the degradation efficiency (DE) of orange II using different samples as the photocatalysts. And the reaction rate constants in different catalyst systems were calculated by the pseudo-first-order model (Fig. 6B) and listed in Table 2. Ti3C2 and CdS show low degradation effect on orange II. Compared with Ti3C2/g-C3N4NS, the CdS-loaded ternary composites show enhanced photocatalytic degradation activity. Among the ternary composites, the CTN-4:1 exhibits the highest photocatalytic activity. The UV–vis spectral variation of orange II solution with different treatment times in the CTN-4:1 system is exhibited in Fig. 6C. It is clear that the typical adsorption peak of orange II at 486 nm decreased rapidly with the time increasing. Simultaneously, the color of the filtered solution changes from deep to colorless, indicating the degradation of orange II. The TOC removal efficiency of orange II in the CTN-4:1 system is about 76 % at 60 min, indicating that orange II can be effectively mineralized (Fig.S6A). After subsequent reactions under identical conditions, the CTN-4:1 still maintains high catalytic performance after four cycles (Fig.S6B). In Fig.S6C, no evident difference is found for the structure of CTN-4:1 before and after the photocatalytic reaction, which indicates excellent reusability and stability.
Analysis Of Photocatalytic Mechanism
Figure 7A and Fig. S7A present the UV–vis absorption spectra of the samples. Ti3C2 shows no obvious absorption edge. Compared with the CdS and Ti3C2/g-C3N4NS, all CdS-loaded ternary composites can cause an appropriate shift to long wavelengths. The (αhν)1/2 versus hv plots of the samples are presented in Fig. 7B and Fig. S7B. The band-gap energy (Eg) values of g-C3N4 and g-C3N4NS are calculated to be 2.87, 2.77eV, respectively. The Eg of obtained g-C3N4NS is lower than that of g-C3N4. And the Eg values of other corresponding samples are summarized in Table 2. Compared with the CdS and Ti3C2/g-C3N4NS, the Eg values of the CdS/Ti3C2/g-C3N4NS composites decrease. Among the ternary composites, the CTN-4:1 has the lowest band-gap energy, which is beneficial improve the absorption of visible light.
Table 2
Reaction rate constant and band-gap energy of the samples
Samples
|
Reaction rate constant (min− 1)
|
Band-gap energy (eV)
|
Ti3C2
CdS
Ti3C2/g-C3N4NS
|
3.8×10− 3
3.9×10− 3
1.3×10− 2
|
0.50
2.27
2.50
|
CTN-2:1
|
1.6×10− 2
|
1.85
|
CTN3:1
|
2.3×10− 2
|
1.82
|
CTN-4:1
|
4.2×10− 2
|
1.76
|
CTN-5:1
|
1.7×10− 2
|
1.89
|
The PL spectra and transient photocurrent responses of the samples were shown in Fig. 8 (A and B) and Fig.S8 (A and B). The CdS/Ti3C2/g-C3N4NS composites show lower PL intensity and higher photocurrent density compared with Ti3C2/g-C3N4NS and CdS. Among all samples, CTN-4:1 exhibits the lowest PL intensity and maximum photocurrent intensity, indicating its higher separation efficiency of the photoinduced e–h+ pairs (Liu et al. 2020). As seen in Fig. 8C, by contrast, CTN-4:1 exhibits the smaller impedance radius, indicating the higher charge transfer and separation ability (He et al. 2019). The formation of heterojunction in ternary composites can promote the transfer and separation of photoinduced carries, which leads to the enhancement of photocatalytic activity
Figure 9A shows the trapping experiments for the CTN-4:1degradation system using BZQ, EDTA, and TBA as the scavengers of •O2−, h+, and •OH species, respectively. The DEs of orange II decrease after the addition of the scavengers, indicating that the •O2−, h+, and •OH species are formed, and •O2− and h+ species are the main active species in the degradation system. Meanwhile, the EPR spectra of DMPO- •O2− in the CdS, Ti3C2/g-C3N4NS and CTN-4:1 degradation systems were determined under vis irradiation. As shown in Fig. 9B, the six characteristic peaks corresponding to DMPO–•O2− are detected in the CTN-4:1 degradation system. The generation efficiency of •O2− in the CTN-4:1 system is slightly higher than those of CdS and Ti3C2/g-C3N4NS. On the other hand, the characteristic signal peaks of DMPO–•OH signals can also be observed for the CTN-4:1 system (Fig.S9). Under vis irradiation, CTN-4:1degradation system can promote the production of active species and effectively degrade orange II.
As shown in Fig. S10A, compared with CdS/g-C3N4NS sample, CTN-4:1catalysis has higher photocatalytic activity for degradation of orange II under the same conditions. The trapping experiments for the CdS/g-C3N4NS system show that both •O2− and h+ are the major active species during the degradation process (Fig.S10B). As presented in Fig. S10C, the generation efficiency of •O2− in the CdS/g-C3N4NS system is lower than that of CTN-4:1. From Fig.S9 we can see that the •OH radicals are basically not produced for the CdS/g-C3N4NS degradation system. The formation of CdS/Ti3C2/g-C3N4NS Z-scheme heterojunction composites using Ti3C2 a charge transport bridge, can improve the redox capacity of photocatalysis and promote the production of the active species.
Figure S11 shows the Mott-Schottky plots of the g-C3N4NS, Ti3C2, and CdS. The slopes of the lines are all positive, implying that the samples belong to n-type semiconductor. The flat potential (Efb) of g-C3N4NS, Ti3C2, and CdS are estimated to be -1.11, -0.22, and − 0.47 V vs.Ag/AgCl (0.22 V vs. NHE), respectively. Normally, conduction band (ECB) of an n-type semiconductor was about 0.1eV more negative than its Efb(Sun et al. 2014; Ma et al. 2018). Thus, the ECB values of g-C3N4NS, Ti3C2, and CdS are estimated to be about − 0.99, -0.10, -0.35 V vs.NHE, respectively. According to the band-gap energies of samples, the valence band (EVB) values of g-C3N4NS, Ti3C2, and CdS are calculated to be 1.78, 0.40, 1.92 V vs.NHE, respectively.
According to the above analysis, the plausible photocatalytic degradation mechanism of the CdS/Ti3C2/g-C3N4NS catalyst is depicted in Scheme 1. Under vis irradiation, e–h+ pairs can be generated on the CdS/Ti3C2/g-C3N4NS Z-scheme composite. The ECB of CdS (− 0.35 V vs. NHE) is more negative than that of Ti3C2 (− 0.10 V vs. NHE), and the photoinduced electrons in CdS can be transferred to Ti3C2. The EVB of g-C3N4NS (1.78V vs. NHE) is higher than that of Ti3C2 (0.40 V vs. NHE). The photogenerated holes in g-C3N4NS can migrate to the Ti3C2. As a result, the photogenerated carriers can be effectively separated by the Z-scheme heterostructure. Meanwhile, the ECB of g-C3N4NS (-0.99 V vs. NHE) is more negative than the potential of superoxide radicals (− 0.046 V) (Kandi et al. 2017), and the dissolved O2 can combine with the electrons in the ECB of g-C3N4NS to produce •O2− in the photodegradation process. In acidic degradation system, some of •O2- radicals may also combine with H+ to produce •OH (Cai et al. 2018). At the same time, the holes are accumulated in the EVB of CdS. The CdS/Ti3C2/ g-C3N4NS degradation system can produce additional active species, resulting in high catalytic degradation efficiency of orange II under vis irradiation.