Efficient spatial charge separation in unique 2D tandem heterojunction Cd_xZn_1−xIn₂S₄–CdS–MoS₂ rendering highly-promoted visible-light-induced H₂ generation

Abstract

Two-dimensional (2D) semiconductor nanostructures have exhibited great prospect as an efficient photocatalyst for solar-to-fuel application. In this work, a unique 2D tandem heterojunction consisting of ultrathin Cd_xZn_1−xIn₂S₄ nanosheets coupled with rectangular CdS flakes and defect-rich MoS₂ few-layered nanosheets was constructed for the first time. Remarkably, the efficient electron transfer channels present in the CdS/Cd_xZn_1−xIn₂S₄ and Cd_xZn_1−xIn₂S₄/MoS₂ 2D tandem heterojunctions facilitate the spatial separation and directional migration of photo-induced charge carriers effectively. Moreover, such 2D tandem heterojunction Cd_xZn_1−xIn₂S₄–CdS–MoS₂ is provided with excellent light harvesting capacity and abundant HER active sites from the defective MoS₂ co-catalyst. These distinct advantages endow the optimized C_0.15ZIS–5C–3M hybrid (5 wt% CdS, 3 wt% MoS₂) with an exceptional photocatalytic H₂ evolution reaction (HER) activity of 27.14 mmol h⁻¹ g⁻¹, approximately 47 times that of pure ZnIn₂S₄ and it is much superior to that of Pt-decorated C_0.15ZIS–5C and most ZnIn₂S₄-based composites reported previously. A high HER apparent quantum yield (AQY) of 19.97% is achieved at λ = 400 nm. In addition, both the cycling and long-term HER measurements evidence the prominent stability of C_0.15ZIS–5C–3M for H₂ production. The results indicated here could pave the way for the exploitation of new 2D heterostructures toward highly-efficient solar conversion and utilization.

---

1. Introduction

Semiconductor photocatalysis for hydrogen (H₂) generation via water-splitting holds enormous promise in ameliorating the worsening energy shortage and environment degradation as it can make direct use of the inexhaustible and clean solar energy.^1–3 So far, numerous semiconductor photocatalysts such as metal oxides,⁴ sulfides,⁵ and nitrides⁶ have been developed systematically. Particularly, the metal sulfides like binary CdS and ternary ZnIn₂S₄ have received tremendous interests as their strong ability for visible-light harvest and appropriate band structures for water reduction. Yet, the photocatalytic H₂ generation efficiency of metal sulfides is still unsatisfactory as a result of the rapid charge recombination along with photocorrosion issues.⁷ To surmount the above problems, several effective approaches, including morphological control,^8,9 heteroatom doping,^10,11 heterojunction construction,^12,13etc. have been demonstrated to afford a significant promotion on the photocatalytic capability.

Recently, two-dimensional (2D) nanostructures have aroused considerable research interests for photocatalysis due to their distinct structural virtues, for instance, the shortened bulk-to-surface transfer distance for photo-excited charge carriers, higher surface areas for the adsorption of reactant species, and more exposed catalytically active sites for redox reactions,^14,15 all of which play crucial roles in improving the photocatalytic activity. Among various 2D metal sulfides, layer-structured ZnIn₂S₄ nanosheets have been widely studied as the matrix of heterostructures for photocatalytic H₂ formation, seeing that the separation and transport of photogenerated charges can be dramatically accelerated by the built-in electric field formed at the heterointerfaces to obtain remarkably boosted photo-activity and stability.¹⁶ Nevertheless, the ZnIn₂S₄ products are usually produced as larger aggregates with lower surface area and active-site amount,^17–19 thus it still remains a big challenge so far to fabricate the free-standing multi-component ZnIn₂S₄-based 2D heterostructures with controllable compositions and interface structures. In addition, serving as the key component of hybrid semiconductors, the co-catalyst has important roles in providing abundant reaction active sites and inhibiting the recombination of charge carriers through the interfacial heterojunctions.²⁰ Hence, optimizing the microstructure of co-catalyst is a vital means to achieve superior photocatalytic ability. Of late, layered transition metal chalcogenides, especially MoS₂, have been extensively utilized as the noble-metal-alternative co-catalyst for H₂ evolution.^21,22 Previous study confirms that,²³ the HER active sites of MoS₂ locate on the edges, while the basal planes are inactive for H₂ formation. As a result, it is desirable to fabricate the MoS₂ nanosheets co-catalyst with an ultrathin thickness and rich structural defects, which can bring about numerous additional active edge sites to facilitate the H₂ production effectively.^24,25 Therefore, the ultrathin defect-rich MoS₂ nanosheets are expected to be a promising co-catalyst for the construction of highly-efficient 2D semiconductor heterostructures with remarkable HER activity, although the relevant investigations are rarely reported to now.

Herein, unique 2D tandem heterojunction Cd_xZn_1−xIn₂S₄–CdS–MoS₂ comprising Cd_xZn_1−xIn₂S₄ ultrathin nanosheets coupled with rectangular CdS flakes and defect-rich MoS₂ few-layered nanosheets was synthesized for the first time. These Cd_xZn_1−xIn₂S₄–CdS–MoS₂ composites are endowed with excellent light harvesting capacity, abundant HER active sites from the defective MoS₂ nanosheets, as well as effective spatial isolation and directional transmission of photogenerated carriers through the efficient electron transfer channels generated in the CdS/Cd_xZn_1−xIn₂S₄ and Cd_xZn_1−xIn₂S₄/MoS₂ 2D tandem heterojunctions. As a result, when irradiated by visible-light (λ > 400 nm), the optimized Cd_xZn_1−xIn₂S₄–CdS–MoS₂ with 5 wt% CdS and 3 wt% MoS₂ (C_0.15ZIS–5C–3M) displays an outstanding HER activity, which is 47 times that of pure ZnIn₂S₄ and much better than that of Pt-loaded C_0.15ZIS–5C and most ZnIn₂S₄-based hybrid photocatalysts ever documented. The present results demonstrate the controllable fabrication of uniform 2D tandem heterojunction Cd_xZn_1−xIn₂S₄–CdS–MoS₂ and the defect-rich MoS₂ few-layered nanosheets as co-catalyst have great potential for highly-efficient photocatalytic H₂ generation. Moreover, the detailed synthesis, characterizations on the morphology and structure, chemical valence states, and physicochemical properties of the prepared photocatalysts are also reported.

2. Experimental section

2.1. Materials

All chemicals used in the present work were analytically pure and supplied by Sinopharm Chemical Reagent, Co., Ltd.

2.2. Synthesis of 2D Cd_xZn_1−xIn₂S₄ ultrathin nanosheets

The Cd_xZn_1−xIn₂S₄ ultrathin nanosheets were prepared via a facile ethanediol-assisted solvothermal method. Typically, calculated amounts of ZnCl₂, CdCl₂·2.5H₂O, and InCl₃·4H₂O for the fabrication of 1 mmol Cd_xZn_1−xIn₂S₄ (labeled as C_xZIS, 0 ≤ x ≤ 1) were first dissolved into 60 mL ethylene glycol under stirring condition. Following this step, 4 mmol thioacetamide was put into the above solution and agitated to obtain a homogeneous mixture. Finally, the solution was sealed in a 100 mL autoclave to carry out the solvothermal reaction at 180 °C for 24 h. After reaction completed, the products were washed using deionized water and ethanol and dried at 70 °C for 5 h under vacuum condition.

2.3. Synthesis of C_0.15ZIS–CdS hybrid nanosheets

Typically, 0.2 g C_0.15ZIS nanosheets were dispersed by sonication into 30 mL deionized water to form a suspension. Then, a certain quantity of CdCl₂·H₂O, C₆H₅Na₃O₇·2H₂O, and CH₄N₂S (mole ratio of 1 : 1.4 : 2) were dissolved into the above suspension under ultrasonic condition. After that, 2 mL NH₃·H₂O was decanted into the foregoing mixture, which was subsequently sealed in a 50 mL autoclave to perform the hydrothermal reaction at 180 °C for 24 h. When reaction finished, the products were rinsed separately by deionized water and ethanol and dried under vacuum condition. The obtained C_0.15ZIS–CdS nano-hybrids are named as C_0.15ZIS–yC, in which y denotes the weight percentage of CdS multiplied by 100.

2.4. Synthesis of defect-rich MoS₂ nanosheets and C_0.15ZIS–5C–MoS₂ hybrid nanosheets

The MoS₂ nanosheets with abundant structural defects were prepared by a hydrothermal method. Generally, appropriate amounts of (NH₄)₆Mo₇O₂₄·4H₂O (1 mmol) and CH₄N₂S (30 mmol) were dissolved into 35 mL deionized water by vigorous stirring to get a transparent solution. After that, the above solution was sealed in an autoclave (50 mL) and heated at 200 °C for 24 h. Finally, the resultant MoS₂ precipitate was washed using deionized water and ethanol and finally dried in a vaccum oven.

The C_0.15ZIS–5C–MoS₂ composites were prepared with a self-assembly strategy. Specifically, 30 mL DMF suspension containing a certain amount of MoS₂ nanosheets was prepared beforehand through an ultrasonic exfoliation treatment for 3 h. Following this step, 0.2 g C_0.15ZIS–5C was mixed with the above suspension by ultrasonication and a 12 h agitation was performed to induce the self-assembly reaction. Finally, the resultant C_0.15ZIS–5C–MoS₂ hybrid nanosheets were washed by centrifugation with ethanol for three times. The C_0.15ZIS–5C–MoS₂ products were abbreviated as C_0.15ZIS–5C–zM, where z equals to the mass fraction of MoS₂ multiplied by 100.

2.5. Synthesis of Pt-loaded C_0.15ZIS–5C hybrid nanosheets via photo-deposition

The Pt-decorated C_0.15ZIS–5C hybrid nanosheets were fabricated via an in situ photocatalytic reduction reaction. Prior to irradiation, 10 mg C_0.15ZIS–5C was dispersed by ultrasonication into 100 mL aqueous solution comprising hole-scavenger. After that, a certain amount of H₂PtCl₆ was added and the mixture was then transferred to the reactor and evacuated for 0.5 h to get rid of the dissolved air. The light source for photo-deposition was provided by a Xe lamp (300 W, PLS-SXE 300D) installed with a cut-off filter. Finally, Pt-loaded C_0.15ZIS–5C hybrid nanosheets can be obtained after the photo-reduction reaction for 1 h.

2.6. Characterization

Powder X-ray diffraction (XRD) signals were examined by an X-ray diffractometer (Rigaku D/max 2500 PC) employing the scanning rate of 5° min⁻¹. Scanning electron microscopy (SEM) measurements were carried out on a ZEISS MERLIN Compact scanning electron microscope. Transmission electron microscopy (TEM), high-resolution TEM (HRTEM), dark-field scanning TEM (STEM), as well as the energy-dispersive X-ray spectroscopy (EDX) results were tested by a transmission electron microscope (JEOL JEM-2100F). Atomic force microscopy (AFM) results were obtained through the use of an atomic force microscope (Bruker MultiMode 8). Brunauer–Emmett–Teller (BET) tests were carried out with an automatic adsorption machine (Micromeritics TriStar II 3020). Electron paramagnetic resonance (EPR) data was obtained by a Bruker A300 spectrometer operating at 113 K. X-ray photoelectron spectroscopy (XPS) spectra were measured using a Thermo Fisher ESCALAB 250 Xi spectrometer. Ultraviolet-visible (UV-vis) absorption was detected on a Shimadzu UV 3600 spectrometer. Photoluminescence (PL) signals were collected with an Edinburgh FLS 980 fluorescence spectrophotometer adopting a 365 nm exciting light.

2.7. Photocatalytic H₂ generation activity

Photocatalytic H₂ production test was carried out with a Labsolar-6A apparatus purchased from Beijing Perfectlight Technology Co., Ltd. The reaction temperature was adjusted as 6 °C by a circulating water-bath device. A PLS-SXE 300D Xe lamp (300 W, 400 nm cutoff filter) was employed to give the required irradiation. For a routine test, 10 mg photocatalyst was dispersed by sonication into 100 mL aqueous solution comprising 10 vol% lactic acid. Then, the reaction solution was vacuumized for 0.5 h to remove the dissolved air. At a specific interval, the generated H₂ was detected by a GC-7900 online gas chromatograph. Besides, 0.35 M Na₂S/0.25 M Na₂SO₃, 10 vol% triethanolamine (TEOA), and 0.375 M ascorbic acid solutions were also used for reaction so as to examine the effect of hole scavenger on the HER property of photocatalyst. HER apparent quantum yield (AQY) under 400 nm irradiation was determined on the basis of the following equation:

2.8. Electrochemical and photoelectrochemical tests

To investigate the electrochemical and photoelectrochemical properties of synthesized photocatalysts, a CHI 660E electrochemical analyzer with a standard three-electrode configuration was used for measurement. Pt plate was used as counter-electrode, and Ag/AgCl was chosen as reference electrode. For working electrode, the preparation of which was implemented according to the procedure bellow: firstly, the photocatalyst ink was drop-casted onto a fluorine-doped tin oxide (FTO) glass slide (contacting area of 2 cm²) and then dried naturally, before calcining at 450 °C for 1 h under flowing N₂. Mott–Schottky curves were measured employing 0.5 M Na₂SO₄ aqueous solution as electrolyte under the testing conditions of 0.01 V amplitude and 500 Hz frequency. Photocurrent response was examined in 10 vol% lactic acid aqueous solution adopting the bias potential of 0.5 V. Electrochemical impedance spectroscopy (EIS) data were collected under open circuit potentials using 0.5 M Na₂SO₄ aqueous solution as electrolyte. The measurement was performed with the testing conditions of 0.005 V amplitude and 0.1–10⁶ Hz frequency.

3. Results and discussion

3.1. Catalyst characterization

The Cd_xZn_1−xIn₂S₄–CdS–MoS₂ 2D heterostructures were synthesized through a three-step strategy (Fig. 1). In the first step, ultrathin Cd_xZn_1−xIn₂S₄ solid solution nanosheets were fabricated through the solvothermal reaction in ethylene glycol. Secondly, by using Cd_xZn_1−xIn₂S₄ nanosheets as substrate, rectangular CdS flakes were grown uniformly onto Cd_xZn_1−xIn₂S₄ nanosheets under hydrothermal conditions to form the Cd_xZn_1−xIn₂S₄–CdS hybrid nanosheets. Finally, the defect-rich MoS₂ nanosheets obtained via a solvothermal method were exfoliated by ultrasonication for 3 h, which were then combined with the Cd_xZn_1−xIn₂S₄–CdS to produce the unique Cd_xZn_1−xIn₂S₄–CdS–MoS₂ ternary hybrids by means of the self-assembly process under stirring condition.

Fig. 1 Diagram illustrating the synthesis of Cd_xZn_1−xIn₂S₄–CdS–MoS₂ composites.

The crystalline structure of products was confirmed by XRD tests. According to Fig. 2a, the diffraction peaks of synthesized ZnIn₂S₄ can be indexed as hexagonal ZnIn₂S₄ (JCPDS no. 65-2023). Interestingly, the diffraction signals of Cd_xZn_1−xIn₂S₄ (x = 0.05, 0.10, 0.15, and 0.20) were successively moved toward lower angle with the increasing Cd ratio (inset), which indicates that the products were Cd-doped ZnIn₂S₄ instead of the mixture of ZnIn₂S₄ and CdIn₂S₄.²⁶ It is well known that, the Cd²⁺ possesses a larger radius (0.97 Å) than that of Zn²⁺ (0.74 Å). Therefore, the lattice expansion happened after the Zn²⁺ sites of ZnIn₂S₄ crystal were occupied by Cd²⁺, resulting in the low-angle-shift of XRD peaks.^27,28 However, when the ZnCl₂ was completely replaced by CdCl₂ for synthesis, the resultant product was CdIn₂S₄ (JCPDS no. 27-0060) mixed with some CdS (JCPDS no. 80-0006) and CdCO₃ (JCPDS no. 52-1547) (Fig. S1†). The XRD patterns of MoS₂, CdS, C_0.15ZIS, as well as their hybrids were displayed in Fig. 2b. The XRD signals of individual CdS and C_0.15ZIS are assigned to the wurtzite CdS and hexagonal ZnIn₂S₄, corresponding to JCPDS no. 65-3414 and no. 65-2023, respectively. Compared to bulk MoS₂ (JCPDS no. 77-1716), the layer-spacing of as-synthesized MoS₂ nanosheets was enlarged as the two peaks of (002) and (004) were shifted to lower angles.²⁹ The increased layer-spacing of MoS₂ nanosheets could contribute to the exposure of more unsaturated S atoms on the edge sites, which serve as the active sites for H₂ evolution. Moreover, the expanded layer-spacing of MoS₂ nanosheets makes their exfoliation much easier to generate abundant active S atoms to promote the HER efficiently. We can see that the XRD patterns of C_0.15ZIS–5C and C_0.15ZIS–5C–3M exhibit the signals originating from C_0.15ZIS component, while the diffraction peaks of CdS and MoS₂ couldn't be observed due to their lower loading amounts.³⁰ By increasing the MoS₂ content to 10 wt%, the (002) peak of MoS₂ appeared in the diffraction pattern of C_0.15ZIS–5C–10M composite.

Fig. 2 (a) XRD patterns of Cd_xZn_1−xIn₂S₄ nanosheets. (b) XRD patterns of MoS₂, CdS, C_0.15ZIS, C_0.15ZIS–5C, C_0.15ZIS–5C–3M, and C_0.15ZIS–5C–10M.

Fig. 3 presents the ultraviolet-visible (UV-vis) absorption spectra of MoS₂, CdS, ZnIn₂S₄, C_0.15ZIS, C_0.15ZIS–5C, and C_0.15ZIS–5C–3M. It is found that the ZnIn₂S₄ displays an optical absorption edge around 462.3 nm, which is moved to 471.9 nm when the C_0.15ZIS is concerned. By comparison to ZnIn₂S₄ and C_0.15ZIS, the light-harvesting capability of CdS is significantly enhanced and the corresponding optical absorption edge is evidently red-shifted to 567.6 nm. Noticeably, the MoS₂ is provided with an intense light-absorption stretching across the ultraviolet-light to visible-light bands. After the C_0.15ZIS was hybridized by CdS and MoS₂, the optical absorption edges of produced C_0.15ZIS–5C and C_0.15ZIS–5C–3M composites are sequentially moved toward longer wavelengths to show a notably improved light-absorption property. Generally, the bandgap of semiconductors can be determined by the Kubelka–Munk equation of (Ahν)ⁿ = K(hν − E_g) (A: absorbance, hν: photon energy, K: constant, E_g: bandgap, n: 0.5 and 2 for indirect-transition and direct-transition semicondutors, respectively).³¹ By employing the Kubelka–Munk method, the bandgaps of CdS, ZnIn₂S₄, C_0.15ZIS, and MoS₂ can be calculated to be 2.29, 2.82, 2.75, and 1.40 eV, respectively (Fig. S2†).

Fig. 3 UV-vis absorption spectra of MoS₂, CdS, ZnIn₂S₄, C_0.15ZIS, C_0.15ZIS–5C, and C_0.15ZIS–5C–3M.

SEM observations were carried out to discern the macroscopic morphology of synthesized ZnIn₂S₄ and C_0.15ZIS. One can find from Fig. 4a that, the synthesized ZnIn₂S₄ is featured by a uniform sheet-like structure. Moreover, as revealed by Fig. 4b, the free-standing sheet-like structure was maintained after the ZnIn₂S₄ was doped by Cd to form C_0.15ZIS. It is noteworthy that, the free-standing ZnIn₂S₄ or C_0.15ZIS nanosheets prepared in this work are quite different from those reported in the literatures,^17,32,33 which are frequently the large-sized aggregates with lower Brunauer–Emmett–Teller (BET) surface area and active-site number, unfavorable for the photocatalytic application. The regular 2D morphology of C_0.15ZIS nanosheets were further confirmed by TEM measurements (Fig. 4c). Additionally, the top and bottom facets are identified as (001) plane according to the fast Fourier transform (FFT) result of the HRTEM graph of a lying C_0.15ZIS nanosheet (Fig. 4d). In addition, the thickness of C_0.15ZIS nanosheets is estimated to be 4.0 nm from the AFM test results shown in Fig. 4e and f. These ultrathin C_0.15ZIS nanosheets are in possession of a large BET surface area (Fig. S3 and Table S1†), which enables them to serve as an ideal matrix to construct novel 2D semiconductor heterostructures with outstanding photocatalytic properties. In addition, elements Cd, Zn, In, and S are uniformly distributed in C_0.15ZIS nanosheets as proved by the STEM and corresponding EDX results in Fig. 4g.

Fig. 4 (a and b) SEM images of (a) ZnIn₂S₄ and (b) C_0.15ZIS nanosheets. (c) TEM image, (d) HRTEM image, (e) AFM image and (f) corresponding height profile, as well as (g) STEM and corresponding EDX elemental mapping results of the C_0.15ZIS nanosheets.

The microstructure of the C_0.15ZIS–CdS–MoS₂ hybrids was studied by TEM tests. Fig. 5a exhibits the typical morphology of the C_0.15ZIS–5C–3M composite, in which many rectangular flakes are uniformly distributed on the surface of bigger nanosheets. On the basis of HRTEM analyses in Fig. 5b, the lattice spacing of 0.33 nm is assigned to (100) plane of C_0.15ZIS nanosheet substrate (JCPDS no. 65-2023). The rectangular flakes display a 0.35 nm lattice spacing, corresponding to (100) plane of CdS (JCPDS no. 65-3414). According to the synthetic procedure of C_0.15ZIS–CdS, pure CdS nanocrystals with irregular shapes were obtained in the absence of C_0.15ZIS nanosheets (Fig. S4†), which confirms the importance of C_0.15ZIS substrate in guiding the formation of rectangular CdS flakes. Meanwhile, the lattice spacing of 0.95 nm is attributed to the expanded layers ((002) plane) of MoS₂ nanosheets (JCPDS no. 77-1716), coinciding well with the XRD results (Fig. 2b). Additionally, there are numerous structural defects (such as lattice distortion and missing, etc.) existed in the MoS₂ few-layered nanosheets as indicated by the cyan dashed circles (Fig. 5b, c and S5†). What's more, the electron paramagnetic resonance (EPR) signal at g = 2.008 demonstrates the presence of S vacancies in the MoS₂ nanosheets (Fig. S6†).³⁴ Similarly, the lattice fringes of rectangular flakes in Fig. 5c own the spacings of 0.31 and 0.35 nm, belonging to the (101) and (100) facets of CdS (JCPDS no. 65-3414), respectively. Besides, the lattice spacing of 0.95 nm also confirms the formation of expanded layers in the MoS₂ component. The above TEM, HRTEM, and EPR data evidence unambiguously the presence of C_0.15ZIS nanosheets, rectangular CdS flakes, and defect-rich MoS₂ nanosheets in the C_0.15ZIS–5C–3M hybrid. Besides, the element distribution of Cd, Zn, In, Mo, and S of C_0.15ZIS–5C–3M was corroborated by the STEM and corresponding EDX results shown in Fig. 5d.

Fig. 5 (a) TEM and (b and c) HRTEM images, (d) dark-field STEM photo and corresponding EDX elemental mapping results of the C_0.15ZIS–5C–3M composite.

XPS investigation was carried out to examine the chemical valence states of surface species for C_0.15ZIS–CdS–MoS₂ composites. Here, we select the C_0.15ZIS–5C–10M as an example for interpretation. As shown in Fig. 6a, elements Cd, Zn, In, Mo, and S were co-existed in C_0.15ZIS–5C–10M composite. The Cd 3d spectrum displays two signals at 404.9 and 411.6 eV (Fig. 6b), which are consistent with that of Cd²⁺ state.³⁵ For Zn 2p spectrum (Fig. 6c), the doublet signals locating at 1021.5–1044.6 eV can be ascribed to the Zn²⁺ species.³⁶ Meanwhile, the In 3d_5/2 and 3d_3/2 signals are observed at 444.7 and 452.2 eV (Fig. 6d), confirming the formation of In³⁺ ions.³⁷ In the Mo 3d spectrum (Fig. 6e), two sets of doublet peaks around 228.1–231.4 eV and 228.9–232.2 eV are assigned to the Mo⁴⁺ and Mo⁵⁺, respectively.^30,38 By comparison to pure C_0.15ZIS and MoS₂ (Fig. S7a–c and e†), the binding energies of Cd 3d, Zn 2p, In 3d, and Mo 3d peaks of C_0.15ZIS–5C–10M are all changed, implying that there exist electron transfer and strong interaction between the C_0.15ZIS, CdS, and MoS₂ components of C_0.15ZIS–5C–10M.³⁹ Regarding to S 2p spectrum, the doublet signals at 161.3–162.4 eV are related to the S²⁻ ligands,⁴⁰ while the doublet peaks at 162.7–163.5 eV match well with the S₂²⁻ species.⁴¹ As documented in the literature,^41,42 the presence of Mo⁵⁺ ions is usually accompanied by the generation of S₂²⁻ species, and both Mo⁵⁺ and S₂²⁻ are the active sites for H₂ evolution. Moreover, the Mo⁵⁺ and S₂²⁻ can also be found in individual MoS₂ (Fig. S7f†), which indicates that these Mo⁵⁺ and S₂²⁻ active sites should originate from the MoS₂ component. Hence, the C_0.15ZIS–CdS–MoS₂ composites decorated with MoS₂ co-catalyst are anticipated to exhibit a superior activity toward photocatalytic H₂ production.

Fig. 6 (a) XPS survey and (b) Cd 3d, (c) Zn 2p, (d) In 3d, (e) Mo 3d, and (f) S 2p spectra of the C_0.15ZIS–5C–10M hybrid.

3.2. Photocatalytic activity and stability

The visible-light-excited (λ > 400 nm) H₂ evolution activity for the prepared samples was studied when lactic acid was used as the hole scavenger. It is seen from Fig. 7a and S8a† that, the MoS₂, CdS, CdIn₂S₄, and ZnIn₂S₄ are all in possession of lower activities toward photocatalytic H₂ evolution. Compared with pristine ZnIn₂S₄, the HER activity of Cd_xZn_1−xIn₂S₄ is increased because of the enhanced light-harvesting (Fig. 3), and the C_0.15ZIS with 15 mol% Cd-doping achieves the maximum HER rate of 2.31 mmol h⁻¹ g⁻¹ (Fig. S9a and b†). The H₂ evolution activity of C_0.20ZIS is lowered in relation to that of C_0.15ZIS, probably due to the smaller BET surface area of the former (Table S1†) that gives rise to less active sites for H₂ evolution. Moreover, after the C_0.15ZIS was hybridized by CdS to form the C_0.15ZIS–CdS composites, the HER performance is notably heightened and the C_0.15ZIS–5C with 5 wt% CdS presents the optimal activity of 7.69 mmol h⁻¹ g⁻¹ (Fig. S9c and d†). The H₂ evolution activity of C_0.15ZIS–CdS is reduced at a higher CdS content, which may result from the excess CdS coverage that diminishes the charge separation efficiency and catalytically active sites on C_0.15ZIS. Interestingly, the HER property of C_0.15ZIS–5C can be further dramatically improved when it was decorated by MoS₂ to produce C_0.15ZIS–5C–MoS₂. Fig. S9e and f† exhibit that, the HER activity of C_0.15ZIS–5C–MoS₂ is elevated with the raising MoS₂ concentration and the highest HER rate of 27.14 mmol h⁻¹ g⁻¹ is achieved at the 3 wt%-MoS₂ (C_0.15ZIS–5C–3M). The apparent quantum yield (AQY) for H₂ evolution of C_0.15ZIS–5C–3M under 420 nm irradiation is calculated as 11.80%, which is raised to 19.97% when the H₂ generation reaction is driven by 400 nm irradiation (Fig. S10†). However, further increasing the MoS₂ content leads to a decreased HER capability. The possible reason is that excess MoS₂ will screen the light-absorption of CdS and C_0.15ZIS, which reduces the amount of photo-induced charge carriers and thus cuts down the H₂ production efficiency accordingly. The H₂ formation rate of C_0.15ZIS–5C–3M is about 47 times that of pure ZnIn₂S₄ and it exceeds that of most known ZnIn₂S₄-based photocatalysts (Table S2†). Additionally, the C_0.15ZIS–5C–3M is far more active than the C_0.15ZIS–5C–3 wt% Pt as well as the bi-component C_0.15ZIS–5C, C_0.15ZIS–3M, and CdS–MoS₂ with the same CdS/MoS₂ ratio as that in C_0.15ZIS–5C–3M (Fig. 7a), suggesting that the excellent photocatalytic activity of C_0.15ZIS–5C–3M results from the synergistic cooperation of the C_0.15ZIS, CdS, and MoS₂ components. On the other hand, the effect of sacrificial agents on the HER capability of C_0.15ZIS–5C–3M was investigated. We can find from Fig. 7b and S8b† that, the H₂ production of C_0.15ZIS–5C–3M is facilitated more effectively in lactic acid solution as compared with that in other solutions containing ascorbic acid, TEOA, or Na₂S/Na₂SO₃. The pH value of lactic acid, ascorbic acid, TEOA, and Na₂S/Na₂SO₃ solutions used for HER is around 1.3, 2.5, 10.8, and 13.4, respectively. Therefore, the higher HER activity of C_0.15ZIS–5C–3M in lactic acid solution could result from the presence of more H⁺ ions for H₂ production due to the smaller pH value of lactic acid solution than the other three solutions. In addition, under long-time irradiation, the steady increase on the evolved H₂ amount of C_0.15ZIS–5C–3M (Fig. 7c) evidences that the C_0.15ZIS–5C–3M possesses favorable stability for sustainable H₂ production from photocatalytic water-splitting. Furthermore, the HER cycling stability of C_0.15ZIS–5C–3M was examined utilizing the same photocatalyst and testing conditions. After five consecutive measurements, no obvious alteration was observed for the HER activity of C_0.15ZIS–5C–3M (Fig. 7d), which also manifests the outstanding photocatalytic stability of C_0.15ZIS–5C–3M for H₂ evolution.

Fig. 7 (a) Photocatalytic H₂ evolution rates of different samples. (b) HER activity of C_0.15ZIS–5C–3M in different sacrificial-agent solutions. (c) Long-term and (d) cycling H₂ generation durability of C_0.15ZIS–5C–3M.

The structure of C_0.15ZIS–5C–MoS₂ after H₂ production reaction was also characterized by XRD, HRTEM, and XPS measurements. As displayed in Fig. S11,† the XRD signals of used C_0.15ZIS–5C–3M were nearly unchanged with respect to the original one, certifying that the crystal structure of C_0.15ZIS–5C–3M was well preserved after catalytic reaction. Moreover, the TEM and HRTEM images in Fig. S12† verify the existence of C_0.15ZIS nanosheets, CdS rectangular flakes, and defect-rich MoS₂ nanosheets in the recovered C_0.15ZIS–5C–3M hybrid. Besides, the XPS spectra in Fig. S13† demonstrate that, the chemical constituents and related valence states of C_0.15ZIS–5C–10M were maintained after HER test. Therefore, the above XRD, HRTEM, and XPS results jointly corroborate the prominent stability of the C_0.15ZIS–5C–MoS₂ composites toward visible-light-driven H₂ evolution.

3.3. Photocatalytic mechanism

The excellent photocatalytic activity of C_0.15ZIS–5C–3M composite could be associated with its superior charge separation capability, which was verified by the photocurrent, EIS, and PL measurements. It is seen from Fig. 8a, the C_0.15ZIS–5C–3M displays a much higher photocurrent density than the MoS₂, CdS, and C_0.15ZIS, suggesting that the charge recombination is effectively impeded by C_0.15ZIS–5C–3M as compared with the latter three.⁴³ Moreover, the EIS Nyquist profiles in Fig. 8b exhibit that, the arc diameter of C_0.15ZIS–5C–3M is reduced in relation to that of CdS, MoS₂, and C_0.15ZIS. It is well known that,⁴⁴ a smaller Nyquist arc diameter is indicative of a lowered charge-transfer resistance. Thus, the charge-transfer resistance of C_0.15ZIS–5C–3M is decreased in contrast with that of CdS, MoS₂, and C_0.15ZIS, which originates from the promoted transfer and separation of photogenerated charges, conforming to the photocurrent results. In addition, the PL spectra of CdS, C_0.15ZIS, and C_0.15ZIS–5C–3M were tested as shown in Fig. 8c. All these three samples present the signals around 470 and 520 nm, corresponding to the excitonic and trap-state emissions, respectively.^30,45 By comparison to CdS and C_0.15ZIS, the PL emission of C_0.15ZIS–5C–3M was notably quenched, manifesting that the latter has a better capability for depressing the charge recombination,⁴⁶ which is in agreement with the photocurrent response and EIS analyses. Therefore, the enhanced charge transfer and separation of C_0.15ZIS–5C–3M is responsible for its outstanding performance toward photocatalytic reaction.

Fig. 8 (a) Transient photocurrent responses, (b) EIS Nyquist curves, and (c) PL spectra of MoS₂, CdS, C_0.15ZIS, and C_0.15ZIS–5C–3M. (d) Mott–Schottky plots of MoS₂, CdS, and C_0.15ZIS. (e) Schematic energy band alignments of MoS₂, CdS, and C_0.15ZIS.

To probe into the origin for the heightened charge separation of C_0.15ZIS–CdS–MoS₂ composites, the electronic band structures of C_0.15ZIS, CdS, and MoS₂ were investigated. The Mott–Schottky curves of C_0.15ZIS, CdS, and MoS₂ were collected using three different testing frequencies (Fig. 8d and S14†), and the corresponding flat-band potentials can be determined by the tangent intersection with horizontal axis as −0.48, −0.76, and −0.01 V vs. RHE, respectively. Moreover, the positive slopes of Mott–Schottky curves indicate the C_0.15ZIS, CdS, and MoS₂ are n-type semiconductors.⁴⁷ It is known that, the conduction band minimum (CBM) of n-type semiconductors is 0.1 V more negative in relation to their corresponding flat-band potential.⁴⁸ Hence, the CBM potential of C_0.15ZIS, CdS, and MoS₂ are estimated as −0.58, −0.86, and −0.11 V vs. RHE, respectively. Combining the CBM positions of C_0.15ZIS, CdS, and MoS₂ with their bandgaps (Fig. S2†), the corresponding band alignments were established. As shown in Fig. 8e, the CBM level of C_0.15ZIS is higher than that of MoS₂ but lower than that of CdS, whilst the valence band maximum (VBM) level for C_0.15ZIS shifts downward with respect to MoS₂ and CdS. Therefore, after visible-light excitation, the photogenerated electrons will first flow from the CB of CdS to the CB of C_0.15ZIS, and subsequently to the CB of MoS₂. Meanwhile, photo-holes in the valence band (VB) of C_0.15ZIS could migrate to the VB of MoS₂ and CdS due to the existed potential difference. Because of the above charge-transfer processes, the recombination of photo-induced charge carriers can be effectively hindered, leading to significantly improved photocatalytic efficiency. Besides, the stronger visible-light-absorption abilities of CdS and MoS₂ (Fig. 3) help to enhance the light-harvesting of C_0.15ZIS–CdS–MoS₂. What's more, there exist abundant HER active sites in the defect-rich MoS₂ nanosheets (Fig. 5b, c, 6, and S5†). Benefiting from the above advantages, the C_0.15ZIS–CdS–MoS₂ hybrids demonstrate an outstanding photocatalytic property toward visible-light-induced H₂ evolution.

4. Conclusions

In summary, unique C_0.15ZIS–CdS–MoS₂ 2D tandem heterojunction photocatalyst was fabricated for the first time, in which the C_0.15ZIS ultrathin nanosheets were coupled by CdS rectangular flakes and defect-rich MoS₂ few-layered nanosheets. Detailed characterizations reveal that this type of 2D tandem heterojunction is critical to the highly-improved HER performance. Benefiting from the presence of efficient electron-transfer channels in the CdS/Cd_xZn_1−xIn₂S₄ and Cd_xZn_1−xIn₂S₄/MoS₂ 2D tandem heterojunctions, the spatial separation and directional migration of photo-excited charge carriers are facilitated effectively as confirmed by the PL, photocurrent response, and EIS results. Meanwhile, the excellent light harvesting and abundant HER active sites are also very conducive to the exceptional photocatalytic activity of Cd_xZn_1−xIn₂S₄–CdS–MoS₂ hybrid. The optimized C_0.15ZIS–5C–3M hybrid (5 wt% CdS and 3 wt% MoS₂) shows the maximum HER rate of 27.14 mmol h⁻¹ g⁻¹, approximately 47 times that of pure ZnIn₂S₄ and it is much superior to that of Pt-loaded C_0.15ZIS–5C and the most known ZnIn₂S₄-based photocatalysts. The apparent quantum yield (AQY) for H₂ production reaches up to 19.97% at λ = 400 nm. In addition, the C_0.15ZIS–5C–3M is in possession of outstanding cycling and long-term H₂ production stabilities under visible-light irradiation. The findings reported here could inspire the rational design of novel 2D semiconductor heterostructures to facilitate the efficient solar conversion and utilization.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors thank the financial support from the National Natural Science Foundation of China (51802170, 51772162, 21801150), the Natural Science Foundation of Shandong Province (ZR2019MB001, ZR2019JQ14), the Youth Innovation and Technology Foundation of Shandong Higher Education Institutions, China (2019KJC004), the Taishan Scholar Project of Shandong Province (ts201712047), the Special Fund Project to Guide Development of Local Science and Technology by Central Government, and the Taishan Scholar Program of Advantage and Characteristic Discipline Team of Eco-Chemical Process and Technology.

Notes and references

1. S. B. Wang, Y. Wang, S. L. Zhang, S. Q. Zang and X. W. Lou, Adv. Mater., 2019, 31, 1903404.
2. Z. Wang, C. Li and K. Domen, Chem. Soc. Rev., 2019, 48, 2109.
3. J. A. Nasir, Z. U. Rehman, S. N. A. Shah, A. Khan, I. S. Butler and C. R. A. Catlow, J. Mater. Chem. A, 2020, 8, 20752.
4. Y. Z. Wei, J. Y. Wang, R. B. Yu, J. W. Wan and D. Wang, Angew. Chem., Int. Ed., 2019, 58, 1422.
5. Y. J. Yuan, D. Q. Chen, Z. T. Yu and Z. G. Zou, J. Mater. Chem. A, 2018, 6, 11606.
6. M. Xiao, B. Luo, M. Lyu, S. C. Wang and L. Z. Wang, Adv. Energy Mater., 2018, 8, 1701605.
7. Y. Xia, B. Cheng, J. J. Fan, J. G. Yu and G. Liu, Small, 2019, 15, 1902459.
8. Y. Z. Wei, N. L. Yang, K. K. Huang, J. W. Wan, F. F. You, R. B. Yu, S. H. Feng and D. Wang, Adv. Mater., 2020, 7, 2002556.
9. J. Xiong, J. Di and H. M. Li, J. Mater. Chem. A, 2020, 8, 12928.
10. R. Shi, H. F. Ye, F. Liang, Z. Wang, K. Li, Y. X. Weng, Z. S. Lin, W. F. Fu, C. M. Che and Y. Chen, Adv. Mater., 2018, 30, 1705941.
11. S. Li, L. J. Zhang, T. F. Jiang, L. P. Chen, Y. H. Lin, D. J. Wang and T. F. Xie, Chem.–Eur. J., 2014, 20, 311.
12. S. Chandrasekaran, L. Yao, L. B. Deng, C. Bowen, Y. Zhang, S. Chen, Z. Q. Lin, F. Peng and P. X. Zhang, Chem. Soc. Rev., 2019, 48, 4178.
13. X. F. Ning and G. X. Lu, Nanoscale, 2020, 12, 1213.
14. Z. Y. Sun, N. Talreja, H. C. Tao, J. Texter, M. Muhler, J. Strunk and J. F. Chen, Angew. Chem., Int. Ed., 2018, 57, 7610–7627.
15. J. Di, J. Xiong, H. M. Li and Z. Liu, Adv. Mater., 2018, 30, 1704548.
16. Y. P. Yuan, L. W. Ruan, J. Barber, S. C. J. Loo and C. Xue, Energy Environ. Sci., 2014, 7, 3934–3951.
17. Y. Q. He, H. Rao, K. P. Song, J. X. Li, Y. Yu, Y. Lou, C. G. Li, Y. Han, Z. Shi and S. H. Feng, Adv. Funct. Mater., 2019, 29, 1905153.
18. S. S. Li, D. S. Dai, L. Ge, Y. Q. Gao, C. C. Han and N. Xiao, Dalton Trans., 2017, 46, 10620.
19. G. Ma, C. Q. Shang, M. L. Jin, L. L. Shui, Q. G. Meng, Y. G. Zhang, Z. Zhang, H. Liao, M. Li, Z. H. Chen, M. Z. Yuan, X. Wang, C. Y. Wang and G. F. Zhou, J. Mater. Chem. C, 2020, 8, 2693.
20. X. F. Zhou, Y. X. Fang, X. Cai, S. S. Zhang, S. Y. Yang, H. Q. Wang, X. H. Zhong and Y. P. Fang, ACS Appl. Mater. Interfaces, 2020, 12, 20579.
21. K. Chang, X. Hai and J. H. Ye, Adv. Energy Mater., 2016, 6, 1502555.
22. X. Zong, H. J. Yan, G. P. Wu, G. J. Ma, F. Y. Wen, L. Wang and C. Li, J. Am. Chem. Soc., 2008, 130, 7176.
23. T. F. Jaramillo, K. P. Jørgensen, J. Bonde, J. H. Nielsen, S. Horch and I. Chorkendorff, Science, 2007, 317, 100.
24. J. Z. Chen, X. J. Wu, L. S. Yin, B. Li, X. Hong, Z. X. Fan, B. Chen, C. Xue and H. Zhang, Angew. Chem., Int. Ed., 2015, 54, 1210–1214.
25. J. F. Xie, H. Zhang, S. Li, R. X. Wang, X. Sun, M. Zhou, J. F. Zhou, X. W. Lou and Y. Xie, Adv. Mater., 2013, 25, 5807–5813.
26. F. Yang, N. N. Yan, S. Huang, Q. Sun, L. Z. Zhang and Y. Yu, J. Phys. Chem. C, 2012, 116, 9078.
27. X. N. Xing, Q. Zhang, Z. Huang, Z. J. Lu, J. B. Zhang, H. Q. Li, H. B. Zeng and T. Y. Zhai, Small, 2016, 12, 874.
28. H. F. Ye, R. Shi, X. Yang, W. F. Fu and Y. Chen, Appl. Catal., B, 2018, 233, 70.
29. J. F. Xie, J. J. Zhang, S. Li, F. Grote, X. D. Zhang, H. Zhang, R. X. Wang, Y. Lei, B. C. Pan and Y. Xie, J. Am. Chem. Soc., 2013, 135, 17881.
30. X. L. Yin, L. L. Li, W. J. Jiang, Y. Zhang, X. Zhang, L. J. Wan and J. S. Hu, ACS Appl. Mater. Interfaces, 2016, 8, 15258.
31. M. Nowak, B. Kauch and P. Szperlich, Rev. Sci. Instrum., 2009, 80, 046107.
32. Y. Z. Wang, D. Chen, L. S. Qin, J. H. Liang and Y. X. Huang, Phys. Chem. Chem. Phys., 2019, 21, 25484.
33. Z. Y. Zhang, K. C. Liu, Z. Q. Feng, Y. N. Bao and B. Dong, Sci. Rep., 2016, 6, 19221.
34. C. Du, Q. Zhang, Z. Y. Lin, B. Yan, C. X. Xia and G. W. Yang, Appl. Catal., B, 2019, 248, 193.
35. M. Y. Xing, B. C. Qiu, M. M. Du, Q. H. Zhu, L. Z. Wang and J. L. Zhang, Adv. Funct. Mater., 2017, 27, 1702624.
36. M. Zhou, S. B. Wang, P. J. Yang, Z. S. Luo, R. S. Yuan, A. M. Asiri, M. Wakeel and X. C. Wang, Chem.–Eur. J., 2018, 24, 18529.
37. S. J. Peng, P. N. Zhu, V. Thavasi, S. G. Mhaisalkar and S. Ramakrishna, Nanoscale, 2011, 3, 2602.
38. H. F. Lin, B. W. Sun, H. Wang, Q. Q. Ruan, Y. L. Geng, Y. Y. Li, J. K. Wu, W. J. Wang, J. Liu and X. Wang, Small, 2019, 15, 1804115.
39. S. M. Zhu, Y. N. Zhang, X. J. Qian, X. X. Wang and W. Y. Su, Appl. Surf. Sci., 2020, 504, 144396.
40. L. Wei, Y. J. Chen, Y. P. Lin, H. S. Wu, R. S. Yuan and Z. H. Li, Appl. Catal., B, 2014, 144, 521.
41. J. H. Xiong, Y. H. Liu, D. K. Wang, S. J. Liang, W. M. Wu and L. Wu, J. Mater. Chem. A, 2015, 3, 12631.
42. Y. H. Chang, R. D. Nikam, C. T. Lin, J. K. Huang, C. C. Tseng, C. L. Hsu, C. C. Cheng, C. Y. Su, L. J. Li and D. H. C. Chua, ACS Appl. Mater. Interfaces, 2014, 6, 17679.
43. D. Q. Zeng, L. Xiao, W. J. Ong, P. Y. Wu, H. F. Zheng, Y. Z. Chen and D. L. Peng, ChemSusChem, 2017, 10, 4624.
44. S. B. Wang, B. Y. Guan and X. W. D. Lou, J. Am. Chem. Soc., 2018, 140, 5037.
45. T. Simon, N. Bouchonville, M. J. Berr, A. Vaneski, A. Adrović, D. Volbers, R. Wyrwich, M. Döblinger, A. S. Susha, A. L. Rogach, F. Jäckel, J. K. Stolarczyk and J. Feldmann, Nat. Mater., 2014, 13, 1013.
46. Q. M. Sun, N. Wang, J. H. Yu and J. C. Yu, Adv. Mater., 2018, 30, 1804368.
47. F. K. Meng, J. T. Li, S. K. Cushing, M. J. Zhi and N. Q. Wu, J. Am. Chem. Soc., 2013, 135, 10286.
48. C. C. Ling, X. J. Ye, J. H. Zhang, J. F. Zhang, S. J. Zhang, S. G. Meng, X. L. Fu and S. F. Chen, Sci. Rep., 2017, 7, 28174428.

---

Footnotes

† Electronic supplementary information (ESI) available: Bandgaps, N₂ adsorption–desorption isotherms, EPR signal, Mott–Schottky curves, tables summarizing the BET surface areas and reported HER data, as well as additional XRD patterns, XPS spectra, TEM images, and HER activities. See DOI: 10.1039/d0ta10564c

‡ Equal contribution to this work.

---