Nowadays, a green and clean energy source is urgently sought owing to insufficient fossil fuel reserves and pollution problems [1-4]. Hydrogen (H2) is an extremely clean energy source, and the only substance that burns in air is water vapor. Furthermore, solar energy is an inexhaustible source of energy. Therefore, photocatalytic H2 evolution is a potential research field [5-9]. To date, many semiconductors such as metal selenides [10-14], metal sulfides [15-19], carbon nitrides [20, 21], and metal oxides [22-26] have been reported for photocatalytic H2 production. It is worth noting that CdS is widely studied owing to its suitable band gap (about 2.4 eV) and band edge, which result in the material exhibiting good performances under visible light irradiation. However, a single layer of CdS displays a high carrier recombination rate under light irradiation [27], which results in serious photo-corrosion. In order to solve this problem of CdS, many methods have been developed, including the construction of heterojunctions with other semiconductors [28-30], formation of solid solutions [31-35].
Many isomorphous crystals produce homogeneous and variable composition solid solutions. When a wide band gap semiconductor and a narrow band gap semiconductor form a solid solution [36, 37], a novel photocatalyst with a continuously varying band gap can be prepared. In recent years, solid solutions have been extensively studied because of their excellent performances under visible light irradiation [38-40]. Zinc sulfide (ZnS) exhibits good photocatalytic activity and stability owing to its high conduction band energy position [41, 42]. However, ZnS has a wide band gap (of about 3.5 eV) and can only be excited with ultraviolet light. Therefore, introduction of Zn ions into CdS to form Zn1–xCdxS solid solutions can combine the photostability of ZnS and the photocatalytic activity of CdS. More importantly, the band structure can be adjusted by changing the molar ratio of Zn to Cd.
In recent years, organic-inorganic hybrid materials have played a huge role in photocatalysis. As a small organic molecule, diethylenetriamine (DETA) can be protonated by solvothermal reaction with water under high pressures to form positively charged amine ions [43]. The protonated amine ions form an organic-inorganic hybrid material by coordination with anions. Our group has synthesized Zn1–xCdxS-DETA with different band gaps by a simple one-step solvothermal method [44]. The organic-inorganic hybrid material can not only adjust the morphology, but also improve their photocatalytic performance. Therefore, solid solutions have great research prospects in photocatalysis. Although the single solid solution improved the photocatalytic performance, its stability decreased after repeated cycles of experiments. In order to overcome this shortcoming, some researchers have previously constructed heterojunctions with other semiconductors, such as CdSe/ZnxCd1–xS [45], α-Fe2O3/Zn0.4Cd0.6S [46], and ZnCdS/CdS [47].
Graphite carbon nitride (g-C3N4), as a two-dimensional material, is widely studied because it is cheap, non-toxic, and displays good photostability [48-50]. Metal-free g-C3N4 can be excited with visible light because of its suitable band gap (about 2.83 eV), and its conduction band position is sufficiently negative (–1.28 eV) [51-54]. Based on the above points, g-C3N4 has great research prospects in the field of photocatalytic H2 production [55, 56]. The shortcoming is that the photogenerated carriers of single g-C3N4 can easily recombine, which hinders its practical application. Some strategies such as doping other elements [57, 58], loading cocatalysts [59-61], tuning the morphology, and forming heterojunctions with other semiconductors have been developed to address this issue [62-66]. We obtained porous g-C3N4 (Pg-C3N4) by thermal decomposition of thiourea and urea. Compared with that of g-C3N4, Pg-C3N4, with a porous structure, exhibits a larger specific surface area. In addition, these pores provide internal channels that accelerate the carrier migration. This offers the possibility of practical application of Pg-C3N4.
In our work, we successfully developed a heterojunction between Pg-C3N4 and Zn0.2Cd0.8S-DETA by an in-situ solvothermal method. The H2 evolution activity of Pg-C3N4/ Zn0.2Cd0.8S-DETA composite was also studied. The results show that the Pg-C3N4/Zn0.2Cd0.8S-DETA composites display enhanced H2 production activities and stabilities, compared to those of other catalysts. Among them, 15%Pg-C3N4/ Zn0.2Cd0.8S-DETA shows the best photocatalytic activity and stability. Specifically, it maintains its excellent activity after seven cycles. The improvement in the photocatalytic activity is attributed to the formation of a step-scheme heterojunction between Pg-C3N4 and Zn0.2Cd0.8S-DETA.
Urea, zinc chloride (ZnCl2), sulfourea, sodium sulfide (Na2S), sublimed sulfur (S), chloroplatinic acid (H2PtCl6), and DETA were obtained from Sinopharm (P.R. China). Absolute ethyl alcohol, sodium sulfate (Na2SO4), polyethylene glycol, cadmium chloride (CdCl2·2.5H2O), Nafion, and sodium sulfite (Na2SO3) were provided by Shanghai Chemical Reagent Co. Ltd (P.R. China). Deionized water (DW; 18.25 MΩ) was used in the experiments.
First, urea and thiourea were mixed in mortar at the ratio of 3:1 and continuously ground. Thereafter, the mixture was placed in a muffle furnace and heated to 550 ℃ for 2 h. Then, the sample was naturally cooled to room temperature, and finally, Pg-C3N4 was obtained.
X%Pg-C3N4/Zn0.2Cd0.8S-DETA (X = 5, 10, and 15) composites were synthesized by an in-situ growth method. First, different grades of Pg-C3N4 were added to the reactor, followed by the addition of 0.0489 g ZnCl2, 0.328 g CdCl2, 24 mL DETA, and 12 mL DW. Thereafter, 0.256 g S was added to the mixed liquid and stirred well for 3 h; then, the reaction kettle was placed in an oven and heated at 80 ℃ for 48 h. After cooling, the composite was washed several times with DW, and the Pg-C3N4/Zn0.2Cd0.8S-DETA system was obtained by freeze drying.
The crystal structures of Pg-C3N4, Pg-C3N4/Zn0.2Cd0.8S-DETA system, and Zn0.2Cd0.8S-DETA were investigated by XRD (Rigaku D/MAX 24000). TEM (JEM-2100 electron microscope) was used to determine the particle sizes and structures of the samples. The chemical compositions of the samples were determined by XPS (Thermo ESCALAB 250) and Fourier transform infrared spectroscopy (FT-IR NICOLET 6700). Moreover, UV-vis diffuse reflectance spectroscopy (DRS) measurements were performed by using a PerkinElmer Lambda 950 UV-vis spectrophotometer. The PL spectra of Pg-C3N4, 15%Pg-C3N4/ Zn0.2Cd0.8S-DETA, and Zn0.2Cd0.8S-DETA were acquired by FLS920 combined fluorescence life time measurements. The photoelectrochemical measurements were performed on a Shanghai Chenhua CHI-660D electrochemical system. The electrolyte solution was 1.0 mol/L Na2SO4. In addition, 0.05 g of the catalyst samples was mixed with 50 μL 5% Nafion and 0.5 mL ethyl alcohol to form a slurry. The slurry was injected into a 1.0 cm2 ITO conductive glass electrode and dried for 30 min at 60 ℃.
The photocatalytic H2 production experiment was carried out in a 250 mL flask with three necks. First, 50 mg of a sample, 100 mL of 0.25 mol/L Na2SO3, and a solution mixed with 0.35 mol/L Na2S and 300 μL H2PtCl6 (0.6 wt%) were placed in the flask with three necks. Then, the mixture was dispersed by ultrasound for 30 min and stirred for another 30 min. The system was bubbled with nitrogen for 30 min under dark conditions. Finally, a 300 W Xe lamp with a 420 nm cut-off filter was used for the irradiation. The distance between the Xe lamp and the reactor was 5 cm. The amount of H2 produced by illumination was measured by gas chromatography (GC-7900).
Fig. 1 shows the synthetic process of 15%Pg-C3N4/Zn0.2Cd0.8S-DETA. First, a certain amount of urea and thiourea were mixed in mortar for grinding. The ground mixture was placed in a crucible, which was then transferred to the muffle furnace and annealed at 550 ℃ for two hours to produce Pg-C3N4. Thereafter, a certain amount of Pg-C3N4 was added to polytetrafluoroethylene liner, and specific amounts of DETA, sublimed S, cadmium chloride, and zinc chloride were added. After the mixture was stirred in a Teflon liner for 3 h, the liner was transferred to a stainless steel reactor and placed in an oven for 48 h at a temperature of 80 ℃, and finally, 15%Pg-C3N4/Zn0.2Cd0.8S-DETA was obtained.
In order to further investigate the phases of the synthesized photocatalysts, we performed XRD on all the photocatalysts. The XRD patterns of Pg-C3N4 and X%Pg-C3N4/Zn0.2Cd0.8S-DETA are shown in Fig. 2. Furthermore, the phases of Zn0.2Cd0.8S-DETA and Pg-C3N4 have been identified by previous work, and are shown in Fig. 2. It is easy to see that 5%Pg-C3N4/ Zn0.2Cd0.8S-DETA, 15%Pg-C3N4/Zn0.2Cd0.8S-DETA, and 25%Pg- C3N4/Zn0.2Cd0.8S-DETA all exhibit the peaks of Zn0.2Cd0.8S- DETA, which indicate that all the composites contain Zn0.2Cd0.8S-DETA. In addition, the main (002) peak of X%Pg-C3N4/Zn0.2Cd0.8S-DETA gradually grows and resembles that of Pg-C3N4 with the increase in X value, which indicates that Pg-C3N4 is present in the composite. From this, it can be concluded that X%Pg-C3N4/Zn0.2Cd0.8S-DETA contains both Pg-C3N4 and Zn0.2Cd0.8S-DETA.
To further explore the micromorphology of each photocatalyst, we also investigated the samples by means of TEM, HRTEM, EDS mapping, and SEM. As can be seen clearly in Fig. 3a, Pg-C3N4 is an ultrathin porous nanosheet with a hole of 20–50 nm. In Fig. 3b, Zn0.2Cd0.8S-DETA exhibits a nanoflower structure with a size of about 300 nm, and the petals of the nanoflowers exist in an ultrathin state. It can be clearly seen from Fig. 3c that the surface of the ultrathin and porous Pg-C3N4 is filled with Zn0.2Cd0.8S-DETA nanoflowers. It can also be seen from Fig. 3d that the spacings of Pg-C3N4 and Zn0.2Cd0.8S-DETA are 0.33 and 0.35 nm, respectively. Additionally, in Fig. 3d, the red curve on the left side clearly shows the heterojunction formed between Pg-C3N4 and Zn0.2Cd0.8S-DETA. This indicates that the 15%Pg-C3N4/Zn0.2Cd0.8S-DETA material we synthesized is a result of the combination of Pg-C3N4 and Zn0.2Cd0.8S-DETA, and not a mixture of two pure substances. Fig. 3e is a SEM image of 15%Pg-C3N4/Zn0.2Cd0.8S-DETA. Fig. 3f shows the EDS pattern of 15%Pg-C3N4/Zn0.2Cd0.8S-DETA. The EDS pattern shows that the composites we synthesized contain only C, N, O, Zn, Cd, and S elements, and no other impurities. The elemental maps are shown in Fig. 3g–3l. It is obvious that the 15%Pg-C3N4/Zn0.2Cd0.8S-DETA composite contains C, N, O, Zn, Cd and S elements and that the elements are evenly distributed.
The XPS patterns are shown in Fig. 4. Fig. 4a reveals the measured spectra of Zn0.2Cd0.8S-DETA, Pg-C3N4, and 15%Pg-C3N4/Zn0.2Cd0.8S-DETA. Zn0.2Cd0.8S-DETA shows the presence of C 1s, O 1s, N 1s, Zn 2p, Cd 3d, and S 2p elements without other impurities. Pg-C3N4 reveals the presence of C 1s, N 1s, and O 1s elements, and no other elements were detected. Among them, the O element found in Pg-C3N4 may have originated from the oxygen in the air. 15%Pg-C3N4/ Zn0.2Cd0.8S-DETA shows the presence of C 1s, O 1s, N 1s, Zn 2p, Cd 3d, and S 2p elements, and no other impurities were detected, which indicate the purity of the substance. High-resolution XPS C 1s patterns are shown in Fig. 4b. 15%Pg-C3N4/Zn0.2Cd0.8S-DETA displays two main peaks, which are due to the inclusion of Pg-C3N4 [67]. However, the peak on the right side is larger, which is attributed to the high content of Zn0.2Cd0.8S-DETA. Moreover, the main peak of 15%Pg-C3N4/ Zn0.2Cd0.8S-DETA was shifted relative to that of Pg-C3N4. Fig. 4c shows the high-resolution XPS N 1s spectra. Among them, 15%Pg-C3N4/Zn0.2Cd0.8S-DETA exhibits a larger peak on the left side and a smaller peak on the right side, which is due to the higher content of Zn0.2Cd0.8S-DETA. However, the peak intensity on the right side of 15%Pg-C3N4/Zn0.2Cd0.8S-DETA is higher than that of Zn0.2Cd0.8S-DETA, which is ascribed to the small amount of Pg-C3N4 in the composite [68]. In addition, the main peak of the composite reveals a different degree of shift compared with that of the main peak of the pure substance. From the changes in the C 1s and N 1s binding energies, it can be seen that the current density of Pg-C3N4 is increased, and it is concluded that electrons flow from Zn0.2Cd0.8S-DETA to Pg-C3N4 [69]. Fig. 4d shows a high-resolution XPS Zn 2p spectra. The two main peaks of 15%Pg-C3N4/Zn0.2Cd0.8S-DETA are shifted by about 0.27 eV relative to those of Zn0.2Cd0.8S-DETA. Fig. 4e is high-resolution XPS Cd 3d spectra. The two main peaks of 15%Pg-C3N4/Zn0.2Cd0.8S-DETA are shifted by about 0.06 eV relative to those of Zn0.2Cd0.8S-DETA. As shown in Fig. 4f, the S 2p pattern can be fitted with two main peaks, and the two main peaks of 15%Pg-C3N4/Zn0.2Cd0.8S-DETA are shifted by about 0.22 eV with respect to those of Zn0.2Cd0.8S-DETA. This further proves the conclusion made above. In Fig. 4b–4f, the main peaks of the 15%Pg-C3N4/Zn0.2Cd0.8S-DETA composite reveal different degrees of shift relative to those of pure Pg-C3N4 or Zn0.2Cd0.8S-DETA, which suggest that the composite we synthesized is not a simple mixture of two pure substances.
In order to investigate the light absorptions of the pure and composite materials, we conducted UV-vis DRS studies. As shown in Fig. 5a, the absorption edges of Zn0.2Cd0.8S-DETA and Pg-C3N4 are about 521 and 449 nm, respectively. A reduction in the light absorption at 320–380 nm can be observed in all the prepared samples. The reason is that the instrument is changing the light. The corresponding band gaps are 2.45 and 2.85 eV, respectively (Fig. 5b). In Fig. 5a, the absorption edge of 5% Pg-C3N4/Zn0.2Cd0.8S-DETA is similar to that of Zn0.2Cd0.8S-DETA. This is due to the low content of Pg-C3N4, which will not significantly affect the optical absorption properties of Zn0.2Cd0.8S-DETA. However, with the increase in Pg-C3N4 content, the absorption edge of Pg-C3N4/Zn0.2Cd0.8S-DETA composite shifts to blue wavelengths, which is attributed to the low absorption capacity of Pg-C3N4. Fig. 5b presents linear transformations of the Pg-C3N4 and Zn0.2Cd0.8S-DETA absorption curves. As can be seen from Fig. 5b, the band gaps of Pg-C3N4 and Zn 0.2Cd0.8S-DETA are 2.83 and 2.48 eV, respectively. The conduction band (CB) and valence band (VB) energy levels are given by
Here, the value of Ec is 4.5 eV, which is the energy of free electrons in the hydrogen scale. X is the absolute electronegativity, which is 4.64 and 5.21 eV for Pg-C3N4 and Zn0.2Cd0.8S-DETA, respectively. According to calculations, the CB energy levels of Pg-C3N4 and Zn0.2Cd0.8S-DETA are –1.28 and –0.53 eV, respectively, and the VB energy levels of Pg-C3N4 and Zn0.2Cd0.8S-DETA are 1.55 and 1.95 eV, respectively.
The N2 adsorption-desorption curves of the prepared samples are shown in Fig. 6a. They correspond to type-IV, according to the classification of Brunauer-Deming-Deming Teller. This indicates the presence of mesopores in the prepared samples. The Brunauer-Emmett-Teller surface areas (SBET) of the prepared samples are displayed in Fig. 6b. It can be seen that Zn0.2Cd0.8S-DETA exhibits a larger SBET (150.3 m2 g–1) than Pg-C3N4 (52.7 m2 g–1). As the content of Pg-C3N4 increases, the SBET of the Pg-C3N4/Zn0.2Cd0.8S-DETA composite decreases.
The rates of photocatalytic H2 production of X%Pg-C3N4/Zn0.2Cd0.8S-DETA (X = 5, 15, and 25) composite, CdS-DETA, and Pg-C3N4 are presented in Fig. 7. The photocatalytic performance of Zn0.2Cd0.8S-DETA is better than that of pure CdS-DETA, which is attributed to the doping of Zn2+ that promotes the photocatalytic activity. However, the photocatalytic activity of pure Pg-C3N4 is very low. It can be clearly seen in Fig. 7 that the H2 production rate of the 15%Pg-C3N4/Zn0.2Cd0.8S-DETA system is 6.69 mmol g–1 h–1, with an apparent quantum efficiency of 17.8%, which is much higher than those of pure Zn0.2Cd0.8S-DETA and Pg-C3N4, which indicates that the addition of appropriate amounts of Pg-C3N4 can significantly promote the photocatalytic activity of Zn0.2Cd0.8S-DETA. The main reason is that the addition of Pg-C3N4 can effectively promote the separation of the photogenerated electrons and holes in Zn0.2Cd0.8S-DETA, thus improving the photocatalytic efficiency of the photocatalyst. However, the photocatalytic activity of 25%Pg-C3N4/Zn0.2Cd0.8S-DETA is much lower than those of Zn0.2Cd0.8S-DETA and CdS-DETA. This suggests that excess Pg-C3N4 inhibits the photocatalytic activity of Zn0.2Cd0.8S-DETA. Therefore, an appropriate amount of Pg-C3N4 can effectively improve the activity of a photocatalyst, which presents a framework for increasing the activities of single photocatalysts.
In order to further explore the stabilities of the materials, we carried out seven cycles of 21 h experiments on 15%Pg-C3N4/Zn0.2Cd0.8S-DETA and Zn0.2Cd0.8S-DETA. As shown in Fig. 8a, Zn0.2Cd0.8S-DETA maintains a good photocatalytic activity over the first three cycles under visible light irradiation, but its photocatalytic activity significantly decreases from the fourth cycle onwards, reaching only 67.7% of the initial value after the seventh cycle. Nevertheless, 15%Pg-C3N4/Zn0.2Cd0.8S-DETA maintains a high photocatalytic activity during the first six cycles, which decreases slightly after the seventh cycle. In addition, Fig. 8b shows the XRD patterns of the recycled and original samples. It can be seen that the main peaks of 15%Pg-C3N4/Zn0.2Cd0.8S-DETA display no obvious shifts after the cycling. This indicates that the addition of an appropriate amount of C3N4 can not only improve the photocatalytic activity of Zn0.2Cd0.8S-DETA, but also significantly enhance its photocatalytic stability.
To investigate the improvement in the carrier separation efficiency of 15%Pg-C3N4/Zn0.2Cd0.8S-DETA, we conducted a photocurrent-time test. As shown in Fig. 9, the light is turned on every 50 s. All the samples show enhanced current intensities when illuminated. The greater the magnitude of the photocurrent, the faster is the charge carrier separation and transmission in 15%Pg-C3N4/Zn0.2Cd0.8S. Therefore, it can be seen that 15%Pg-C3N4/Zn0.2Cd0.8S-DETA exhibits the highest electron-hole pair separation and transmission efficiency. This result is consistent with its H2 production performance.
Based on the above discussion, a possible charge transfer mechanism in photocatalysis is proposed. As shown in Fig. 10, the possible mechanism involves a step-scheme that is based on the change in the binding energy of each element in the XPS spectra (Fig. 4). When sunlight is absorbed, the photogenerated electrons are excited from the VBs of Pg-C3N4 and Zn 0.2Cd0.8S-DETA to their CBs, which leave holes in their VBs. After Pg-C3N4 and Zn0.2Cd0.8S-DETA are combined, an internal electric field is formed between the two[70]. The unwanted electrons in the CB of Zn0.2Cd0.8S-DETA recombine with the holes in the VB of Pg-C3N4 owing to the presence of an internal electric field [71]. Then, the electrons in the CB of Pg-C3N4 are transferred to Pt to undergo a reduction reaction with H+ [72], and the holes in the VB of Zn0.2Cd0.8S-DETA are oxidized with S2– and SO32–. In addition, S2– and SO32– can provide S for the catalyst and reduce photoetching of Zn0.2Cd0.8S-DETA. Through the S-scheme mechanism, the electron-hole pairs are separated in space to endow the Pg-C3N4/Zn0.2Cd0.8S-DETA composites with a strong redox capacity. The space-separated holes and electrons undergo oxidation and reduction reactions, respectively. The electron-hole pairs are not enriched in the CB of Zn0.2Cd0.8S-DETA and VB of Pg-C3N4, thereby reducing the photoetching of Zn0.2Cd0.8S-DETA. Therefore, the prepared catalyst can effectively suppress the recombination of electron-hole pairs, which results in excellent photocatalytic performance and photostability.
We have successfully prepared a step-scheme Pg-C3N4/ Zn0.2Cd0.8S-DETA system by a solvothermal method for photocatalytic H2 evolution. 15%Pg-C3N4/Zn0.2Cd0.8S-DETA exhibits the best H2 evolution performance (6.69 mmol g–1 h–1). In addition, 15%Pg-C3N4/Zn0.2Cd0.8S-DETA remains highly active after seven cycles of experiments. The improvement in the photocatalytic performance and photostability is attributed to the formation of a step-scheme heterojunction. The formation of the step-scheme heterojunction promotes the separation of carriers and the transmission efficiency and reduces the probability of recombination. We have provided possible strategies for solving the energy problems.
2020,
Vol. 41 Issue (1): 41-49 DOI: S1872-2067(19)63389-9 
PDF
2020,
Vol. 41 Issue (1): 41-49 DOI: S1872-2067(19)63389-9 PDF