Flower-like MoS2 microspheres compounded with irregular CdS pyramid heterojunctions: highly efficient and stable photocatalysts for hydrogen production from water

An irregular CdS pyramid/flower-like MoS2 microsphere composite photocatalyst was successfully synthesized using a simple one-step hydrothermal method. The as-prepared samples were characterized by X-ray diffraction, X-ray photoelectron spectroscopy, scanning electron microscopy, ultraviolet visible absorption spectroscopy, fluorescence spectroscopy and photoelectrochemical tests. The composite photocatalysts showed superior photocatalytic activities for hydrogen evolution from water under visible light irradiation (λ ≥ 420 nm) with an extremely high apparent quantum yield (AQY = 64.8%) at 420 nm. To our knowledge, this value is the highest reported efficiency value for CdS/MoS2 photocatalysts. Further detailed characterization revealed that the special structure for some CdS pyramid structures dispersed in the MoS2 microsphere structures and surrounded by MoS2 nanosheets led to the photogenerated electrons migrating from the conduction band of different faces of the CdS pyramid to the conduction band of different MoS2 nanosheets while photogenerated holes remained in the CdS pyramid structures, which greatly promoted the separation of photogenerated electrons and holes, improving the photoactivity of the CdS/MoS2 catalyst. The catalyst also exhibited perfect stability, and the photoactivity displayed no significant degradation during continuous hydrogen production over nearly 70 h.


Introduction
Hydrogen is a type of clean and nonpolluting renewable energy that has attracted great interest and has the potential to solve the global energy crisis and reduce environmental pollution. An ideal approach for preparing hydrogen using solar energy to generate photocatalytic water splitting could convert solar energy to clean and storable hydrogen energy. The key issue to achieve hydrogen production from solar photocatalytic water splitting is to develop semiconductor catalysts with high activity and stability. 1-8 Among many different kinds of photocatalysts, sulde photocatalysts have more matching band gaps and higher visible light absorption efficiencies for hydrogen production, which have attracted much attention and have shown potential in industrial applications. [9][10][11][12][13][14][15][16][17] CdS, in particular, is the most widely studied sulde photocatalyst because of its small band gap (approximately 2.3 eV). 14 Moreover, CdS loaded by some noble metals, including Pt, Pd, Ru and so on, showed a much higher photogenerated charge efficiency, which greatly improved the CdS photocatalytic hydrogen production activity. [18][19][20] CdS loaded with 0.3 wt% Pt and 0.13 wt% PdS exhibited the best photocatalytic hydrogen production activity with an apparent quantum yield of 64.8% at 420 nm. 18,19 However, considering the cost, some inexpensive metals and compounds as cocatalysts have attracted increasing interest. [21][22][23][24][25] Investigations have indicated that MoS 2 -loaded CdS photocatalysts can also realize high photocatalytic activity.
Many highly active MoS 2 -loaded CdS photocatalysts have been reported in the literature. [26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43] Zhao et al. reported a platinum-free 1D/2D CdS/MoS 2 photocatalyst, and their experimental results showed that the highest hydrogen production rate of 1.79 mmol g À1 h À1 was obtained when the reaction ratio of CdS to MoS 2 was 0.3. 41 Yin et al. reported noble-metal-free CdS@MoS 2 core-shell nanoheterostructures. 6 wt% MoS 2loaded CdS exhibited the best photocatalytic H 2 evolution performance thus far, with a rate of 62.55 mmol g À1 h À1 . 34 Jiang et al. reported a berry-shaped (b)-CdS/MoS 2 photocatalyst, and 1 wt% MoS 2 -loaded CdS exhibited the best photocatalytic H 2 evolution performance thus far, with a rate of 63.59 mmol g À1 h À1 . 40 Reddy et al. successfully synthesized a few-layered black phosphorus/MoS 2 nanohybrid as a promising cocatalyst, 8 wt% of which loaded on CdS nanorods showed a much higher hydrogen production rate of 183.24 mmol g À1 h À1 . 35 Ultrasmall cobalt nanocrystals embedded in 2D-MoS 2 nanosheets as efficient cocatalysts for solar-driven hydrogen production were reported by Lee et al. Thus, a Co-MoS 2 cocatalyst loaded on CdS nanorods showed a very high H 2 production rate (275 mmol g À1 h À1 ) when the mass fraction of MoS 2 loaded with 1 wt% Co was 6%, which was the most active MoS 2 -loaded CdS photocatalyst for hydrogen production reported in the literature. 33 Through analysis of the abovementioned literature, only a small amount of MoS 2 loaded on CdS exhibited a high photocatalytic activity for hydrogen production. Here, we report a kind of MoS 2 /CdS photocatalyst with ower-like MoS 2 microspheres compounded with irregular CdS pyramids. The special structures led to the CdS/MoS 2 photocatalyst exhibiting a superior separation efficiency for photogenerated electrons and holes. When the molar ratio of MoS 2 to CdS was 1 : 1, the MoS 2 /CdS photocatalyst showed the highest hydrogen production rate of 394 mmol g À1 h À1 , and the apparent quantum yield reached 64.8% at 420 nm. To our knowledge, this value is the highest efficiency ever reported for MoS 2 -modied CdS photocatalysts.

Synthesis
All chemicals were of analytical grade and were used as received without any further purication. Cadmium acetate (Cd(CH 3 -COO) 2  Five proper amounts of ammonium molybdate (the molar of 0.01, 0.0071, 0.005, 0.0042 and 0.0033 mol) were dissolved in deionized water (70 mL) in Teon tubes (100 mL), and then, appropriate amounts of Cd(CH 3 COO) 2 $2H 2 O (the molar ratios of Cd to Mo were 0.5, 0.7, 1, 1.2 and 1.5) and excessive thiourea were added to the solution, with stirring for several minutes. The Teon tubes were transferred to autoclaves, which were sealed and heated at 200 C for 24 h. The resulting precipitates collected by centrifugal separation were washed with deionized water and absolute ethanol several times and then dried under vacuum for 24 h at 80 C. The catalysts prepared by this method were denoted as CdS/  2 -i (molar ratios of Cd to Mo, i ¼ 0.5, 0.7, 1, 1.2 and 1.5). "CdS/ MoS 2 " was "CdS/MoS 2 -1" in "3. Results and discussion".
Pure CdS and MoS 2 were also prepared using the same process described above.
Based on CdS/MoS 2 -1, a Pt-loaded CdS/MoS 2 -1 catalyst was prepared by an in situ photoreduction method using a Xe lamp (300 W) equipped with a 420 nm cutoff lter, as follows: a certain amount of CdS/MoS 2 -1 was added to a 200 mL aqueous solution containing 10 vol% lactic acid and 0.25 M Na 2 SO 3 /0.35 M Na 2 S. A certain amount of chloroplatinic acid (2 wt% of the CdS/MoS 2 -1 quality) was added to the two aqueous solutions. Then, the Ptloaded CdS/MoS 2 -1 catalyst was obtained aer the aqueous solution was photoreduced for 1 h, denoted as Pt/CdS/MoS 2 -1.
Pure CdS and MoS 2 were mechanically mixed, and then a Ptloaded catalyst was prepared by the same method, denoted as Pt/CdS + MoS 2 .

Characterization
The X-ray diffraction (XRD) patterns of the as-prepared photocatalysts were obtained on a PANalytical X'pert MPD Pro X-ray diffractometer equipped using Cu-Ka irradiation. Scanning electron microscopy (SEM) images were obtained using a JSM-7800F-type eld emission scanning electron microscope. X-ray photoelectron spectroscopy (XPS) measurements were obtained on an Axis Ultra Kratos (UK) multifunctional spectrometer with monochromatic Al Ka radiation. The ultraviolet visible (UV-vis) absorption spectra were measured on a HITACHI U-4100 spectrophotometer. Fluorescence spectroscopy was performed using a PTI QM-4 uorescence spectrophotometer.

Photocatalytic hydrogen production
Photocatalytic reactions for hydrogen production were performed in a side irradiation Pyrex cell. A total of 0.01 g of catalyst powder (CdS/MoS 2 -i, Pt/CdS/MoS 2 -I or Pt/CdS + MoS 2 ) was added to an aqueous solution (200 mL) containing 10 vol% lactate acid or 0.35 M Na 2 S/0.25 M Na 2 SO 3 as electron donors. Nitrogen was purged in the cell to remove oxygen before irradiation. The solution was irradiated by visible light through a Xe lamp (300 W) equipped with a 420 nm cutoff lter. The amount of hydrogen was determined using TCD gas chromatography (NaX zeolite column, TCD detector, N 2 as carrier gas). The apparent quantum yield (AQY) was calculated according to the equation reported in the literature. 11 3. Results and discussion Fig. 1 shows SEM images of the samples. As shown in the images ( Fig. 1a and b), the morphology of the samples presented differently sized irregular pyramid structures with particle sizes of 10-300 nm (Fig. 1a) and ower-like microspheres with particle sizes of 4-7 mm (Fig. 1b). The ower-like microsphere structures were self-assembled by a plurality of nanosheets with a thickness of 5.94 nm (Fig. 1c-e). Some irregular pyramid structures were surrounded by nanosheets with microsphere structures, and different surfaces of irregular pyramid structures directly contacted nanosheets, as shown in Fig. 1f. The irregular pyramid structures and ower-like microspheres were presumed to be CdS and MoS 2 , and in addition, some CdS pyramid structures were dispersed in the MoS 2 microsphere structures.
To further prove the above inferred conclusion, SEM mapping of the CdS/MoS 2 catalyst was performed, as shown in Fig. 2. As seen, S was distributed throughout the entire catalyst region (Fig. 2b), demonstrating that Mo and Cd were in the form of MoS 2 and CdS in the catalyst. The distribution of Mo (Fig. 2d) was more intensive in the distribution region of the microsphere structures (Fig. 2a), illustrating that MoS 2 formed microsphere structures in the catalyst. The content of Cd was very low, as shown in Fig. 2c, and was distributed in the MoS 2 microsphere structures (Fig. 3c). Some box samples in Fig. 2c were used for element analysis, and the results are shown in Table 1. It can be seen from the table that the microsphere structure was MoS 2 , and a small amount of CdS pyramid structures was dispersed in the MoS 2 microsphere structure. Fig. 3 shows the XRD patterns of CdS/MoS 2 -1. As shown from the patterns, the diffraction peaks of the hexagonal phase CdS (JCPDF no.     electrons in CdS decreased and that in MoS 2 increased, resulting in an increase and decrease in the binding energy for CdS and MoS 2 , respectively; thus, it was inferred that photoelectrons transferred from CdS to MoS 2 , which achieved efficient separation of photogenerated charges between CdS and MoS 2 , improving the photocatalytic activity for hydrogen production. The photoluminescence spectra of pure CdS and CdS/MoS 2 were also measured and are shown in Fig. 6. The spectra present emission peaks at 475-550 nm, which were related to the recombination process of electrons and holes in the semiconductor. Compared with pure CdS, the CdS/MoS 2 peak showed a much lower intensity, revealing that CdS/MoS 2 exhibited a superior efficiency of separation for photogenerated electrons and holes, which improved the photocatalytic activity for hydrogen production. Photocurrent experiments were also performed to further illustrate the charge transfer behavior for CdS/MoS 2 -1 and pure CdS, as shown in Fig. 7. The photoelectrodes for photoelectrochemical test samples were prepared by electrophoretic    . 7 The current-potential characteristics of as-prepared pure CdS and CdS/MoS 2 -1. Test condition: A 300 W Xe lamp coupled with an AM 1.5 filter was used as the light source; photocurrent density was measured in the 0.5 M Na 2 SO 3 aqueous solution as the electrolyte; Hg/Hg 2 Cl 2 was as the reference electrode. deposition (EPD) onto uorine-doped tin oxide (FTO)-coated glass substrates. As shown in Fig. 7, the photocurrent density of the samples increased gradually with increasing applied potential. Compared with pure CdS, CdS/MoS 2 -1 exhibited an extremely higher photocurrent density, indicating that CdS/ MoS 2 -1 showed more efficient charge separation than pure CdS, which improved the photocatalytic activity for hydrogen production.
The amount and average rate of photocatalytic hydrogen production during 5 h irradiation for CdS/MoS 2 -i (molar ratios of Cd to Mo, i ¼ 0.5, 0.7, 1, 1.2 and 1.5) are shown in Fig. 8. With increasing molar ratio, the amount and average rate of hydrogen production increased gradually and then reached maximum values of 19.7 mmol and 394 mmol h À1 g À1 , respectively, for CdS/MoS 2 -1. The amount and average rate decreased as the molar ratio continued to increase. The average rates of photocatalytic hydrogen production for mechanically mixed CdS and MoS 2 were also measured. The experimental results of hydrogen production showed a low average rate of 19 mmol h À1 g À1 for mechanically mixed CdS-and MoS 2-loaded Pt. It was revealed that the interface between CdS and MoS 2 in CdS/MoS 2 contributed to the separation of photogenerated electrons and holes.   Through all of the characterization analyses, it was known that the CdS/MoS 2 photocatalyst exhibited a superior efficiency of separation for photogenerated electrons and holes, which was attributed to the special structure of some CdS pyramid structures dispersed in the MoS 2 microsphere structures and surrounded by MoS 2 nanosheets. The different faces of the CdS pyramid structures contacted the surface of the MoS 2 nanosheets to form heterojunctions. Through the heterojunctions, photogenerated electrons could be transferred from the different faces of CdS pyramid structures to different MoS 2 nanosheets, while photogenerated holes remained in CdS pyramid structures. Many highly active dangling bonds existed at the edge of the nanosheets, which could form chemical bonds with H. The ability to bond with H provided the best medium for the convenient adsorption and desorption of H on the surface of the photocatalyst, accelerating the photochemical reaction rate of the photocatalyst interface. Therefore, the photogenerated electrons could continue to migrate to the edge of the MoS 2 nanosheets with highly active dangling bonds and then react with H + to form H 2 , while the photogenerated holes were consumed by the lactic acid on the surface of the CdS pyramid structures, as shown in Fig. 9. Based on the above analysis, it was revealed that photogenerated electrons migrated from the conduction band of different faces of the CdS pyramid to the conduction band of different MoS 2 nanosheets, while photogenerated holes remained in the CdS pyramid structures, which greatly promoted the separation of photogenerated electrons and holes, improving the photoactivity of the CdS/ MoS 2 catalyst, as shown in Fig. 10.
It is known that the photogenerated electrons could be transferred from the CdS to MoS 2 or Pt, while photogenerated holes remained in CdS, resulting in a superior efficiency of separation for photogenerated electrons and holes in the CdS catalyst, and improving the photocatalytic hydrogen production efficiency for CdS catalyst. The photoactivity of CdS was greatly improved by coloaded Pt, mainly because Pt could capture photogenerated electrons from CdS, further promoting the separation of photogenerated carriers. To provide additional evidence for the migration path of the photogenerated electrons between CdS and MoS 2 in the CdS/MoS 2 catalyst, the hydrogen production activity for Pt-loaded CdS/MoS 2 -1 was studied, as shown in Fig. 11a. The results showed that the average rate of photocatalytic hydrogen production for CdS/MoS 2 -1 was much   higher than that of Pt-loaded CdS/MoS 2 . It is known that the Ptloaded CdS catalyst exhibits excellent photoactivity in the sacricial agent system of Na 2 SO 3 /Na 2 S. To eliminate this effect, the activity of hydrogen production for Pt-loaded CdS/MoS 2 -1 in a 0.25 M Na 2 SO 3 /0.35 M Na 2 S solution was studied, as shown in Fig. 11b. The results showed that the average rate of CdS/MoS 2 -1 was still higher than that of Pt-loaded CdS/MoS 2 -1. According to the above experimental results, photogenerated electrons migrated from the conduction band of CdS to the conduction band of MoS 2 . Pt could capture photogenerated electrons of CdS, leading to a decrease in photogenerated electron migration to MoS 2 conduction. This, in turn, resulted in a decrease in the hydrogen production activity for CdS/MoS 2 . This nding was consistent with the experimental results.
The stability of CdS/MoS 2 -1 in a lactic acid solution was studied, as shown in Fig. 12. CdS/MoS 2 -1 exhibited perfect stability, the photoactivity displayed no signicant degradation during continuous hydrogen production over nearly 70 h, and the photoactivity of CdS/MoS 2 -1 was reduced by only 1.6 mmol aer 72 h of photocatalytic reaction.

Conclusions
In summary, an irregular CdS pyramid/ower-like MoS 2 microsphere composite photocatalyst was successfully synthesized, achieving a hydrogen evolution rate of 394 mmol g À1 h À1 with an extremely high apparent quantum yield (AQY ¼ 64.8%) at 420 nm. To our knowledge, this value is the highest efficiency ever reported for MoS 2 -modied CdS photocatalysts. Because of the special structure of some CdS pyramid structures dispersed in the MoS 2 microsphere structures and surrounded by MoS 2 nanosheets, the photogenerated electrons migrated from the conduction band of different faces of the CdS pyramid to the conduction band of different MoS 2 nanosheets. Meanwhile, photogenerated holes remained in the CdS pyramid structures, which greatly promoted the separation of photogenerated electrons and holes, improving the photoactivity of the CdS/ MoS 2 catalyst. The catalyst also exhibited perfect stability, and the photoactivity displayed no signicant degradation during continuous hydrogen production over nearly 70 h. This research has some guiding signicance for promoting the study of lowcost and efficient photocatalytic hydrogen production.

Conflicts of interest
There are no conicts to declare.