3.1 Structure and morphology of the materials
3.1.1 The SEM characterization
The type, ratio, and diffusion behavior of the solvent have a large influence on the catalyst morphology[33, 34]. Figure 1 shows the SEM images of the catalyst generated at three-volume ratios of DETA, EtOH, and H2O. As shown in Fig. 1a and d, the CdS-F/0 samples exhibit flower-like structures when DETA and EtOH were used as solvents with a volume ratio of 2: 1, and each flower is constructed of several assembled nanosheets[32]. After adding H2O (VDETA: VEtOH: VH2O=2:1:0.2) to the reaction system, it was found that macropores appeared in the lamellar units forming the flowers, which resulted in the formation of a porous flower micro-nanostructure (Figs. 1b and e). However, net-like CdS micro-nanostructures were observed in the CdS-N/0 samples shown in Figs. 1c and f, without the addition of ethanol (VDETA: VEtOH: VH2O=2:0:1).
3.1.2 X-ray diffraction and X-ray photoelectron spectroscopy analysis
Table 1
Structural properties of CdS nanostructures determined from XRD data.
| Crystal structure | XRD intensity I (002)/ {I (100) + I (101)} |
CdS-F/0 | hexagonal | 1.62 |
CdS-PF/0.2 | hexagonal | 1.53 |
CdS-N/0 | hexagonal | 1.44 |
The crystalline structure of the prepared CdS samples was analyzed by XRD and shown in Fig. 2a. All samples displayed the CdS with hexagonal wurtzite structure(JCPDS No. 77-2306)[32]. The diffraction peaks at 24.8°, 26.5°, 28.2°, 36.6°, 43.7°, 47.8°, and 50.9° correspond to crystal faces (100), (002), (101), (102), (110), (103), and (112), respectively. When the solvents are DETA and EtOH, the CdS-F/0 samples show high intensity in the (002) peak and broad peaks for other XRD lines, which implies higher crystallinity along the c-axis and poorer crystallinity in the x-y plane, resulting in a two-dimensional flower-like morphology [35–37]. Furthermore, as the proportion of water in the solvent gradually increases, the relative strength of the (002) plane to the sum of the (100) and (101) planes decreases, and crystallization occurs in the x-y plane and is preferentially oriented along the c-axis[34]. The calculated results are shown in Table 1 (CdS-F/0 > CdS-PF/0.2 > CdS-N/0), and the morphology of the catalysts realizes the transition from two-dimensional flower to one-dimensional net-like. The peak intensity of the crystals gradually increases during the transition from the flower-like to the net-like morphology. The morphology growth mechanism will be covered in further detail in the growth mechanism section below. The composition and elemental chemical states of CdS-N/0 nanomaterials were further analyzed by XPS. Figure 2b displays the full spectrum of CdS-N/0, no other elements were found, except for the Cd, S, O, and C. The Fig. 2c-d present the high-resolution spectrum of Cd 3d and S2p species. Two significant peaks at 404.86 and 411.65 eV, correspond to Cd 3d5/2 and Cd 3d3/2, respectively[38]. The binding energies can be derived into two peaks (161.37 eV and 162.47 eV) which are assigned to S 2p3/2 and S 2p1/2 [39].
3.1.3 Brunauer–Emmett–Teller measurements
Table 2
Band gap of catalysts with different morphologies. BET surface-area and pore volume (Vpore) and average pore diameter (Dpore), were determined from N2 physisorption.
Catalyst | Band-gap (eV) | SBET(m2·g− 1) | Vpore(cm3·g− 1) | Dpore(nm) |
CdS-F/0 | 2.33 | 43.4 | 0.21 | 18.9 |
CdS-PF/0.2 | 2.31 | 78.2 | 0.27 | 16.2 |
CdS-N/0 | 2.36 | 157.9 | 0.64 | 13.6 |
Figure 3 depicts typical nitrogen adsorption-desorption isotherms and pore size distribution curves for the prepared catalysts with different morphology[40]. All isotherms have a type IV (Brunauer-Deming-Deming-Teller (BDDT) classification) form and are mesoporous materials[41]. It can be observed from Table 2 that the specific surface area of samples CdS-F/0, CdS-PF/0.2, and CdS-N/0 ranges from 43.4 m2g-1 to 157.9 m2g-1. Meanwhile, as shown in Fig. 3a, the isotherm of CdS-N/0 shifted higher in the low P/P0 range (0.2), indicating a larger specific surface area of sample CdS-N/0, compared with CdS-F/0 and CdS-PF/0.2 samples[42]. The pore distributions of the as-synthesized materials covered a wide range from 2 to 100 nm, as shown in Fig. 3b. The CdS-F/0 and CdS-PF/0.2 samples had the highest number of pores with a size of about 5 nm, and the number of pores decreased slowly with the increase of pore size. A large number of distributions in the range of 5–50 nm and 5-100 nm appear in CdS-F/0 and CdS-PF/0.2, respectively. In contrast, the pores of CdS-N/0 are mainly concentrated at around 18 nm and distributed between 2–71 nm. Moreover, combined with the nitrogen adsorption-desorption isotherm, CdS-N/0 samples show hysteresis loops at high relative pressures close to unity, indicating the formation of large mesopores and macropores[42].
3.1.4 Morphology growth mechanism
The possible growth mechanisms for CdS-F/0, CdS-PF/0.2, and CdS-N/0 catalysts are depicted in Fig. 4. For the oxygen-free solvent DETA, the contained N atoms are triple-coordinated and smaller than the C atoms. Therefore, the N atoms in the current solvent can bond with the exposed Cd atoms on the (100) and (010) surfaces, thus controlling the growth of the (100) and (010) surfaces. The number of N3C atoms in the solvent molecule and the length of the atomic chain determine the morphology and dispersion of CdS[42]. However, a single DETA solvent is highly viscous and the CdS samples tend to grow into a massive structure, which is unfavorable for dispersion. Therefore, it is necessary to introduce a mobile co-solvent, water or ethanol, into the solvothermal system[32]. Due to the attraction of N atoms to H+ in the chemical equation, the DETA solution is weakly basic. It can be further concluded that if H2O is present in the mixed solvent thermal system, the attraction of N-Cd decreases due to the equilibrium between N-Cd and N-H+. During the formation of CdS-F/0, due to the absence of water in the solvent, the growth of the (100) and (010) crystallites is maximally inhibited in the presence of only N-Cd, while the orientation of the (002) crystallites reach an optimal level, forming a flower-like morphology consisting of nanosheets grown by C-axis extension. Upon addition of water, N-H+ promotes the growth of some of the (100) and (010) crystalline planes and decreases the orientation of the (002) crystalline planes, which leads to the formation of large holes in the nanosheets due to the growth of the x-y plane thereby leading to the formation of a porous flower morphology. In the absence of ethanol participation, the CdS crystals will preferentially grow along the (002) direction and terminate at the (100) or (010) surface to form a well-dispersed net-like structure.
3.2 Photocatalytic Performance of CdS Nanoarchitectures
Different morphologies of CdS were used for photocatalytic dehydrogenation coupling of ethanol. H2, AA, and 2,3-BDO as the main products were formed during 10h of visible light irradiation using CdS as photocatalysts in 5 vol% ethanol solution. As shown in Fig. 5a, the ethanol conversion of CdS-F/0, CdS-PF/0.2, and CdS-N/0 catalysts were 1.57%, 3.6%, and 4.7% respectively. Figure 5b exhibited the hydrogen production rates of the catalysts, it was found that the hydrogen production amount of all CdS samples gradually increased with the reaction time, and CdS-N/0 has the best photocatalytic hydrogen production rates of about 1.32 mmol g− 1 h− 1. Figure 5c presents the production rates of 2,3-BDO and AA computed using the external standard method. The 2,3-BDO and AA yields of CdS-N/0 with larger specific surface area and better crystallinity were 0.61 mmol g− 1 h− 1 and 1.96 mmol g− 1 h− 1, respectively, which were better than those of the CdS-F/0 and CdS-PF/0.2 samples. The selectivity of 2,3-BDO and AA was analyzed in Fig. 5d. CdS-F/0, CdS-PF/0.2, and CdS-N/0, and exhibit selectivities towards 2,3-BDO at 12%, 27%, and 32%, respectively. This may be due to their different pore size distributions as shown in Fig. 3b. But Hydroxyl radicals are more readily peroxidized to form AA. The selectivity of CdS-F/0, CdS-PF/0.2, and CdS-N/0 for AA was 75%, 62%, and 51%, respectively.
To investigate the impact of water content in solvents on photocatalytic performance, a control experiment was conducted using CdS-N/0 photocatalyst in ethanol solutions with different water content, as shown in Fig. 6. Figures 6a and b illustrate the photocatalytic performance of CdS-N/0 catalysts for the conversion of ethanol at different water contents. The photocatalytic performance of the catalyst is weak in pure ethanol solution. When 5 vol% of water was added, the conversion performance of ethanol was significantly improved, especially the hydrogen production was sharply increased, which was mainly because water as an electron sacrificer was able to consume a large number of electrons, effectively inhibiting the electron-hole pair complexation so that more holes were involved in the oxidation process of ethanol. And the hydroxyl radicals generated by hole oxidation form hydrogen bonds with water, which facilitates the desorption of the generated hydroxyethyl radicals from the catalyst surface and increases the possibility of C-C coupling. When 10 vol% water is introduced, the quantity and structure of hydrogen bonds change, complicating the coupling process and lowering photocatalytic efficiency. The selectivity of the coupling reaction to produce 2,3-BDO in an ethanol solution containing 5 vol% of water is also superior to other ratios of water, as shown in Fig. 6b.
3.3 The light absorption and photoelectrochemical performance of the prepared samples
To better evaluate the photocatalytic efficiency, the optical properties of CdS with different morphologies were investigated using UV-Vis diffuse reflectance spectroscopy. As show in Fig. 7a, the evident and narrow absorption edges show good visible light absorption[32]. Figure 7b shows the converted plots based on the Kubelka–Munk function vs. photon energy, from which the bandgap (Eg) energy of CdS-F/0, CdS-PF/0.2, and CdS-N/0 can be roughly estimated to be 2.33, 2.31, and 2.36 eV, respectively[43]. The band gaps of the sample CdS are all slightly redshifted compared to the bulk CdS (2.42 eV)[44]. Comparison of hydrogen production, 2,3-BDO, and AA yields of CdS samples (Fig. 5) with the optical properties of similar bandgap values (Fig. 7) reveals that nanomaterials with similar bandgaps exhibit different properties, which may be related to the nanomorphology of the catalysts[25].
The photoelectrochemical method was used to evaluate the separation efficiency of photogenerated electrons and holes. Figure 7c-d shows the transient photocurrent density curves and electrochemical impedance spectroscopy (EIS) of the as-prepared CdS-F/0, CdS-PF/0.2, and CdS-N/0 samples. As shown in Fig. 7c, CdS-F/0 exhibits a relatively weak photocurrent response due to the severe recombination of photogenerated electron–hole pairs. CdS-N/0 showed the highest photocurrent response, resulting in excellent light-induced electrical performance. Figure 7d shows that the arc radii of the CdS-N/0 sample is the smallest, indicating the benefits of the Photocatalytic performance[43]. Therefore, the one-dimensional net-like morphology of CdS-N/0 catalysts can promote the separation and transfer of photogenerated charges, which can effectively improve photocatalytic performance.
3.4 Mechanism of photocatalysis
By investigating the effect of water content in the solution on the conversion of ethanol, the hydrogen production rate during the reaction of an ethanol solution with a water content of 5 vol% spiked as compared to the pure ethanol solution, suggesting that the water in the solution consumed most of the photogenerated electrons for the precipitation of hydrogen. CdS catalysts are incapable of generating hydroxyl radical actives. Thus, the actives that play a role in ethanol oxidation are photogenerated vacancies[45]. The αC-H bond in ethanol was oxidized to form ·CH(OH)CH3, and the hydrogen bond formed between the hydroxyl radical and the hydroxyl group of water and the pore structure of the CdS-N/0 catalyst, which was dominated by large mesopores, promoted the desorption of ·CH(OH)CH3 from the surface of the catalyst, which greatly enhanced the possibility of the coupling reaction to produce 2,3-BDO. The hydroxyethyl group which could not be desorbed in time was further oxidized by the holes to other by-products such as acetaldehyde and acetic acid. The reaction process is identical to that shown in Scheme 1.
From the photocatalytic experimental data of the photocatalysts prepared in the above experiments, it can be seen that the photocatalytic performance of the photocatalysts prepared with CdS-N/0 is much higher than that of other morphologies photocatalysts. As far as the light absorption performance of CdS catalysts is concerned, the change in morphology has little effect on the light absorption performance of the catalysts. The excellent photocatalytic performance of CdS-N/0 is mainly attributed to i) the unique one-dimensional structural morphology of CdS-N/0 with a larger specific surface area than other morphologies, which provides more active sites for ethanol conversion; and ii) the better crystallinity and stronger photocurrent intensity of CdS-N/0, which suggests that the photogenerated carriers can be separated better than those of other CdS samples, which is more conducive to the ethanol conversion. Therefore, we conclude that the one-dimensional CdS-N/0 morphology with a larger specific surface area is favorable for photocatalytic ethanol conversion and has the best photocatalytic performance.
At the same time, CdS-N/0 has relatively excellent 2,3-BDO selectivity. From its pore size distribution diagram, it can be concluded that the most available pore size of CdS-F/0 is about 5 nm, and the number of pores decreases slowly with the increase of pore size, with a larger percentage of pores in the 5 nm-50 nm range. Compared with CdS-F/0, the pore distribution in CdS-PF/0.2 is concentrated in the range of 5 nm-100 nm, with a higher number of mesopores and the presence of some macropores. The most available pore size of CdS-N/0 is 18 nm and the percentage of pores from 18 nm to 71 nm is the maximum. We assume that the structure of the CdS-N/0 samples with a larger percentage of large mesopores may offer more possibilities for the formation of 2,3-BDO by coupling reactive[13]. In parallel, we also investigated the number of hydroxyethyl groups produced by CdS samples by electron spin resonance (EPR) spectroscopy using 5,5-dimethyl 1-pyrroline-N-oxide (DMPO) as a spin-trapping agent(Fig. 8). The number of hydroxyethyl radicals produced by the CdS samples was proportional to the selectivity of 2,3-BDO. The correlation between the C-C coupling reaction and the pore size of the catalyst was further demonstrated.
3.5 Modification of photocatalytic performance
Platinum nanoparticles act as reducing co-catalysts that utilize light-generated electrons to reduce H+ and generate H2, thus making it easier to separate and utilize photogenerated carriers[46]. As shown in Fig. 9, the photodeposition of platinum nanoparticles on CdS-N/0 catalysts significantly improves the photocatalytic ethanol conversion performance. The ethanol conversion rate is 16.14%. As shown in Fig. 9a the yield of various products has increased by nearly 3.5 times, with 2,3-BDO yielding 2.64 mmol g− 1 h− 1, AA yielding 6.42 mmol g− 1 h− 1, and H2 yielding 5.45 mmol g− 1 h− 1. As shown in Fig. 9b, the selectivity of the various products remains essentially unchanged compared to CdS-N/0. Figure 9c and d shows the stability of the photocatalytic performance of the Pt/CdS-N/0 catalyst. It can be seen that the yields of the products were basically consistent during the three stability cycle tests. The photocatalytic performance of the catalyst only marginally diminished in the third cycle, which was attributed to the photo corrosion phenomena of the CdS catalyst itself. Figure 9d depicts the product selectivity throughout the catalyst stability test, and it can be seen that the product selectivity remained essentially unchanged during the three cycles of the test.
Based on the photocatalytic properties of the photocatalysts in the previous experiments, it can be determined that the better the crystallinity of the different morphologies of the photocatalysts, the larger the specific surface area, the better the separation of photogenerated carriers, and the better the ethanol conversion performance. The presence of large-sized pores promotes the selectivity of the C-C coupling reaction.