A novel bicomponent Co₃S₄/Co@C cocatalyst on CdS, accelerating charge separation for highly efficient photocatalytic hydrogen evolution

Abstract

In order to realize the dream of the practical application of hydrogen energy from solar energy and water, it is necessary to use efficient and earth-abundant cocatalysts for photocatalytic H₂ evolution. Compared with a one-component cocatalyst, the multicomponent cocatalysts with a synergistic effect can effectively accelerate the separation of photogenerated electron–hole pairs in photocatalysis. Hence, a novel bicomponent Co₃S₄/Co@C cocatalyst on a CdS semiconductor photocatalyst is successfully synthesized by a one-step hydrothermal method. The Co₃S₄/Co-CdS photocatalyst obtained exhibits a highly efficient photocatalytic H₂ evolution rate of 15.17 mmol h⁻¹ g⁻¹, which is 2.8 times higher than that of Pt-CdS. Furthermore, the quantum efficiency of Co₃S₄/Co-CdS reaches a maximum value of 16.80% at 475 nm. The significant improvement in the performance is because of the unique energy level structure of the Co₃S₄/Co-CdS photocatalyst, which has the maximum efficiency of charge separation and migration. This work not only presents a new protocol for constructing bicomponent cocatalysts on a semiconductor structure, but also proves that the bicomponent Co₃S₄/Co@C is a promising and low-cost cocatalyst.

---

Introduction

Hydrogen, which is a highly calorific, clean fuel, can be used to solve the energy crisis and environmental pollution which are caused by the use of traditional nonrenewable fossil fuels.^1–3 Photocatalytic hydrogen production is regarded as a feasible and ideal approach to produce highly purified hydrogen.⁴ One-component photocatalyst exhibits low-efficiency H₂ evolution, because of the high activation potentials required for H₂ production and the rapid recombination of the photogenerated electron–hole pairs. Hence, the multi-component photocatalysts built of semiconductors and cocatalysts are usually designed to achieve efficient photocatalytic hydrogen production.⁵ Cocatalysts can provide the highly active sites for H₂ and O₂ evolution and promote the separation of electron–hole pairs, improving the photocatalytic performance.⁶

The cocatalyst for H₂- and O₂-evolution can enhance both the reliability and activity of the photocatalysts in photocatalytic water splitting based on the following three main factors: (i) compared with a semiconductor, the cocatalyst has a lower overpotential for hydrogen evolution reaction (HER)⁷ and oxygen evolution reaction (OER),^8,9 (ii) the cocatalyst can promote the separation and migration of photogenerated charge carriers at the semiconductor-cocatalyst interface,^10,11 and (iii) the cocatalyst can suppress the photocorrosion and enhance the stability of photocatalyst.¹² The HER system using sacrificial agents has been widely developed for efficient hydrogen production. However, the reported cocatalysts for HER are mostly noble metals, such as Pt, Au and Pd.^13–15 Hence, exploring inexpensive cocatalysts is highly desirable. To date, cost-effective and earth-abundant transitional metals or transition metal sulfides, such as Co, Cu, Co_xS_y and MoS₂,^16–21 have been reported as excellent cocatalysts for catalyzing proton reduction in a photocatalytic reaction. Furthermore, the coupling of metal and a semiconductor results in a Schottky junction at the interface of metal and semiconductor, facilitating the carrier separation and migration, hence dramatically boosting the efficiency of photocatalytic reactions.^22,23 For instance, Kuang et al. reported that CsPbBr₃ coupled with Pd could significantly enhance the photoelectron consumption rate due to the Schottky junction.²⁴ The Schottky barrier will prohibit photogenerated electrons from returning to the semiconductor from the metal, thus, accelerating the electron kinetics of the photocatalytic reaction.²⁵ Compared with the semiconductor photocatalyst, the cocatalyst usually has better electrical conductivity, which is important for the capture and transfer of photogenerated electrons, thus accelerating the carrier separation and migration.²⁶ In conclusion, the cocatalyst not only serves as an electron sink, but also provides earth-abundant and effective proton reduction sites.

The research on cocatalysts is mostly confined to one component, and the exploration of multicomponent cocatalysts is still at a preliminary stage. The synergistic effects induced by multicomponent cocatalysts, such as the efficient charge migration and enhanced photostability, have been reported over the years.^27,28 The synergistic effect at the interface of multicomponent cocatalysts leads to the modification of the electronic structure of different components, which contributes to the electron transfer reaction.²⁹ The large contact interface is critical to obtain the strong coupling effect of the different components.³⁰ Xiong et al. reported that the Pd@Pt cocatalysts significantly enhanced the activity of TiO₂ for photocatalytic water splitting, compared with Pt or Pd, because the multicomponent Pd@Pt cocatalysts prolonged the lifetime of photogenerated electrons.³¹ At present, the research on multicomponent cocatalysts focusses on precious metals. Hence, exploring low-cost and earth-abundant materials as photocatalytic multicomponent cocatalysts is still challenging research.

With these motivations, the bicomponent Co₃S₄/Co@C cocatalyst coupled with a CdS semiconductor was prepared successfully through a one-step hydrothermal method using Co@C as a precursor. The Co₃S₄/Co-CdS heterostructure is well suited for the efficient photocatalytic H₂ evolution due to two primary factors. Firstly, the Co metal can easily accept the photogenerated electrons from the CdS semiconductor, and this one-way transfer prevents the electrons from recombining with holes in the CdS. Secondly, the stepwise down Fermi level structure of Co and Co₃S₄ stimulates the transfer of electrons from Co to Co₃S₄, further promoting the charge separation and migration. Hence, the maximal photocatalytic H₂ evolution rate of Co₃S₄/Co-CdS (15.17 mmol h⁻¹ g⁻¹) is 2.8 times higher than that of Pt-CdS. As the electron sink, the bicomponent Co₃S₄/Co@C cocatalyst not only displays a much better catalytic performance than Pt, but also provides a strategy for the substitution of noble metal cocatalysts.

Experimental section

Preparation of samples

A portion of Co(NO₃)₂·6H₂O (2.99 g), 1.92 g of p-phthalic acid and 2.72 g of triethylene diamine were added to 150 mL of N,N-dimethylformamide under strong agitation. The mixed solution was evaporated at 80 °C until the solvent was completely removed. Then the powders were heated at 80 °C in the oven for 2 h. The as-prepared dry powders were calcined at 900 °C for 1 h under a N₂ atmosphere, with a heating rate of 1.5 °C min⁻¹. The Co@C cocatalyst was finally obtained after cooling down to room temperature.³²

A portion of Cd(CH₃COO)₂·2H₂O (CdAc₂, 0.2665 g) and 0.15 g of thioacetamide (TAA) were put into a 100 ml Teflon-lined stainless-steel autoclave.³³ Next, 1.44–11.52 mg of the as-obtained Co@C samples, corresponding to 1.0%–8.0% mass ratios of Co@C to CdS, were mixed with 50 ml of deionized water with high intensity ultrasonic mixing for 20 min. Then the mixtures were transferred into the Teflon-lined stainless-steel autoclave and heated at 180 °C for 24 h. The as-prepared samples were washed several times with deionized water and ethanol. After drying for 8 h, the Co₃S₄/Co-CdS was obtained. The prepared Co₃S₄/Co-CdS with different mass ratios of Co@C to CdS (1.0, 2.0, 4.0, 5.0, 6.0, and 8.0 wt%) were labelled as SCS1, SCS2, SCS4, SCS5, SCS6 and SCS8, respectively. Under the same experimental procedure, the CdS was prepared without adding Co@C cocatalysts, and the Co₃S₄/Co@C cocatalyst was synthesized using Co@C as a precursor without adding Cd(CH₃COO)₂. Carbon (C) was prepared by a method similar to that of Co@C without adding Co(NO₃)₂·6H₂O. The Co₃S₄@C was also prepared by a method similar to that of Co@C, but the hydrothermal time was extended to 48 h when the metal Co was almost vulcanized into Co₃S₄.

For comparison, the mechanically mixed samples of CdS with Co@C, Co₃S₄/Co@C, Co₃S₄@C or C were prepared by a self-assembly method.³⁴ Typically, 7.20 mg of Co@C, Co₃S₄/Co@C, Co₃S₄@C or C was mixed with 144 mg of CdS, and ground for 40 min. The prepared samples were denoted as Co-CdS, (Co₃S₄/Co)-CdS, Co₃S₄-CdS and C-CdS, respectively. The Pt-CdS was prepared by an in situ photo-deposition method using H₂PtCl₆ aqueous solution (0.01 M).

Material characterizations

The morphologies of the samples were acquired using field-emission scanning electron microscopy (FESEM) on a Philips FEI Quanta 200 FEG instrument. Transmission electron microscopy (TEM) images were obtained on Philips-FEI Talos F200S microscope. X-ray powder diffraction (XRD) patterns were acquired using a Rigaku D/Max 2550 X-ray diffractometer with a Cu Kα radiation source (λ = 1.5418 Å). X-ray photoelectron spectroscopy (XPS) was used to investigate the surface chemical states of the photocatalysts, which were obtained using a Thermo ESCALAB 250 with an Al Kα X-ray source. Ultraviolet photoelectron spectrometry (UPS) was performed using a VG Scienta R4000 analyzer. The UV-vis diffuse reflectance spectra were recorded on a Hitachi U-3010 spectrophotometer. Photoluminescence spectra (PL) were obtained using a Hitachi F-7000 fluorescence spectrophotometer with an excitation wavelength of 360 nm. The SPV spectra were obtained using an Au-Light CEL-SPS 1000 (China). The time-resolved photoluminescence (TRPL) spectra were recorded with an Edinburgh Instruments FLS920 with an excitation wavelength of 380 nm, and the decay curves were fitted by the decay curves of the ExpDec2 function software.

Electrochemical measurements

Photoelectrochemical measurements were carried out on a Gamry, Reference 600+ electrochemical workstation with a three-electrode system, using platinum foil, Ag/AgCl and 0.35 M Na₂S–0.25 M Na₂SO₃ aqueous solution as the counter electrode, reference electrode and electrolyte, respectively. The working electrodes were prepared as follows: 2 mg of catalysts and 30 μL of naphthol were dispersed in 200 μL of ethanol and 770 μL of deionized water, followed by ultrasonic treatment for 20 min. Then 10 μL of the catalyst ink was dropped on the glassy carbon electrode (2.5 mm in diameter). The electrocatalytic activities (HER and OER) were evaluated by linear sweep voltammetry (LSV) at a scan rate of 10 mV s⁻¹, and the NaOH electrolyte was firstly purged with pure N₂ for 1 h. Electrochemical impedance spectroscopy (EIS) were acquired under 10 mV amplitude in the frequency range of 0.1 Hz to 100 kHz in the dark. All of the potentials were calibrated to the RHE.

Photoelectrochemical property measurements

Photoelectrochemical tests were also carried out on a Gamry Reference 600+ electrochemical workstation with the same three-electrode system, using 1 M NaOH aqueous solution as the electrolyte. Typically, the working electrodes were prepared as follows: 1 mg of the as-prepared samples and 20 μL of naphthol were dispersed in 180 μL of ethanol, followed by ultrasonic processing for 20 min. Then 15 μL of the previous solution was uniformly dispersed onto FTO glass with an effective area of 1 cm². Finally, the FTO glass was dried at 60 °C for 1 h. The Mott–Schottky plots were obtained at 1000 Hz. The transient photocurrent responses were measured under the illumination of a CEL-HXF 300, 300 W Xe lamp with a 420 nm cut off filter and bias potential of 0 V vs. Ag/AgCl. Under visible-light illumination, EIS was also tested in the frequency range of 0.01 Hz to 10 kHz.

Photocatalytic activity evaluations

The photocatalytic H₂ evolution tests were performed in a gas-circulating CEL-SPH2N-D9 (Au-Light, China) photocatalytic H₂ evolution system under vacuum with magnetic stirring. A portion (5 mg) of the as-prepared sample was added to 80 ml of 0.35 M Na₂S–0.25 M Na₂SO₃ aqueous solution, followed by ultrasonic treatment for 10 min. The light source was CEL-HXF300, 300 W Xe lamp equipped with 420 nm cut off filter. Photocatalytic H₂ production was determined by a GC7920-TFA (Au-Light, China) on-line gas chromatograph equipped with a thermal conductivity detector (TCD). The apparent quantum efficiency (AQE) under the same conditions of photocatalytic H₂ evolution was carried out at various monochromatic lights, which were obtained using bandpass filters (420, 450, 475, 500, 520 and 550 nm). The AQE at different wavelengths was calculated according to the following equation: AQE (%) = (2 × number of evolved H₂ molecules/number of incident photons) × 100.

Computational methods

The work functions of the cocatalysts were evaluated using the density functional theory (DFT) calculations of the DMol3 software package. The correction, based on Grimme's method, was used for the dispersion interaction.³⁵ The atomic cutoff setting was 4.8 Å for all the elements. The k-point was set to 4 × 4 × 1. Furthermore, the water solvent effect was regarded as a COSMO solvation model with a dielectric constant of 78.54.³⁶ The electrons were spin-polarized in all calculations. Both optimization and band structure calculations were tightly converged.

Results and discussion

The preparation process of the samples is shown in Fig. 1a. The Co₃S₄/Co-CdS photocatalyst was synthesized by a facile one-step hydrothermal method using Co@C as the precursor. The TAA not only reacted with CDAc₂ to form CdS, but also vulcanized Co into Co₃S₄. The TEM and SEM images of Co@C in Fig. 1b show that the Co nanoparticles with sizes of 10–20 nm are coated by the graphitic carbon. The lattice spacing values of 0.20 and 0.33 nm were attributed to the (111) facet of Co and the (002) facet of graphitic carbon, respectively, as shown in Fig. 1c. The EDX element mappings (Fig. S1, ESI†) confirmed that the Co nanoparticles were dispersed homogeneously in the graphitic carbon. The CdS alone existed in the form of nanospheres (Fig. S2a–2c, ESI†), and the lattice spacing values of 0.34 and 0.32 nm corresponded to the (002) facet and the (101) facet of hexagonal phase CdS (Fig. 1d), respectively. When the Co@C cocatalysts were added into the synthetic process, the CdS still kept its the nanosphere structure and hexagonal phase, as shown in Fig. S2d (ESI)† and Fig. 1e. Further observation showed that partial Co nanoparticles were vulcanized into Co₃S₄ (Fig. 1e), therefore, the Co₃S₄/Co@C cocatalysts were successfully obtained during the hydrothermal process. The HR-TEM image shows that the lattice spacing values of SCS5 were 0.20, 0.24, 0.34, 0.32 and 0.25 nm, corresponding to the (111) facet of Co, (400) facet of Co₃S₄,¹⁶ (002) facet, (101) facet and (102) facet of CdS, respectively, further confirming that the Co₃S₄/Co-CdS photocatalyst had been successfully synthesized. In addition, the white frame shown in Fig. 1e indicated the compact interaction between Co and Co₃S₄, that was beneficial for the efficient interfacial charge transfer.³⁴ The STEM image and EDX element mappings of SCS5 in Fig. 1f showed that the Co and Cd elements were distributed in agreement with the S elements, and CdS was combined with the Co₃S₄/Co@C cocatalyst.

Fig. 1 (a) Illustration of the synthesis of Co₃S₄/Co-CdS. (b) TEM and SEM (inset) and (c) HR-TEM images of Co@C. HR-TEM images of (d) CdS and (e) Co₃S₄/Co-CdS. (f) STEM image and the EDX element mapping of SCS5.

The XRD was applied to investigate the crystal structure and phase, and the results are shown in Fig. 2a. The XRD patterns of all the as-prepared samples indicated the existence of a hexagonal wurtzite structure of CdS (JCPDS 41-1049), which was in agreement with the previous HR-TEM results. Furthermore, the unchanging XRD patterns indicated that the formation of the Co₃S₄/Co-CdS heterostructure did not affect the hexagonal wurtzite structure of the CdS. However, no intrinsic XRD peaks of Co, Co₃S₄ or graphitic carbon were observed even though the loading of the cocatalysts was up to 8%, which was probably caused by the high-dispersion of the Co₃S₄/Co@C cocatalyst. The XPS was used further to reveal the chemical valence status and composition of SCS5. The high-resolution XPS spectrum of Cd 3d (Fig. 2b) in the CdS shows two peaks at 405.4 eV and 412.1 eV, which were attributed to Cd 3d_5/2 and Cd 3d_3/2,³⁷ respectively. The S 2p with double peaks (Fig. 2c) was ascribed to the typical metal–sulfur bonds in CdS.^38,39 The XRD and XPS results confirmed that the CdS kept its original structure after coupling with Co₃S₄/Co@C. In the Co 2p high-resolution spectrum (Fig. 2d), a pair of peaks attributed to Co²⁺ ions at 781.89 eV for Co 2p_3/2 and 796.70 eV for Co 2p_1/2 were observed, and another doublet at 784.91 eV and 798.98 eV were ascribed to the Co 2p_3/2 and Co 2p_1/2 of the Co³⁺ ions,^40,41 which indicated the presence of Co₃S₄. The peaks at 787.22 eV and 803.45 eV were satellite signals. Except for the peaks of Co²⁺ and Co³⁺ ions, the spin–orbit doublets belonging to Co⁰ at 787.22 eV for Co 2p_3/2 and 803.45 eV for Co 2p_1/2 were clearly observed,⁴² suggesting the presence of cobalt metal. These results combined with the results of the TEM analyses confirmed that the Co₃S₄/Co-CdS photocatalyst has been successfully synthesized.

Fig. 2 (a) XRD patterns of different samples. The XPS spectra of various elements: (b) Cd 3d, (c) S 2p, (d) Co 2p. (e) Photocatalytic H₂ production rates, error bars represent standard deviation of three tests. (f) Apparent quantum efficiencies of SCS-5 based on different wavelengths and the corresponding UV-visible absorption spectrum. (g) Recyclability of hydrogen evolution tests of SCS-5. (h) UV-vis diffuse reflectance spectra. (i) Electrochemical impedance spectra in the frequency range of 0.01 to 10 kHz under visible light illumination.

The photocatalytic H₂ evolution of the samples were carried out under visible light irradiation (λ > 420 nm), as shown in Fig. 2e. The Co@C cocatalyst had no photocatalytic activity, and the pure CdS displays the low photocatalytic H₂ evolution (0.65 mmol h⁻¹ g⁻¹). The H₂ evolution was substantially enhanced, when the CdS was coupled with the Co₃S₄/Co@C cocatalyst. It is obvious that there was a volcano-type trend between the loading of the Co₃S₄/Co@C cocatalyst and the rate of H₂ evolution. At the optimal loading of 5%, the SCS5 sample had the highest photocatalytic H₂ evolution rate of 15.17 mmol h⁻¹ g⁻¹, which was 23.3 times higher than that of the CdS without the cocatalyst. Furthermore, the SCS5 exhibited a very efficient photocatalytic H₂ evolution rate when compared with other similar composites (Table S1, ESI†). However, by further increasing the loading of the cocatalyst, a deterioration in the photocatalytic performance was observed, because of the covering of the surface-active sites on the CdS by excess cocatalyst. In addition, the determination of the AQE was carried out under different wavelengths (420, 450, 475, 500, 520 and 550 nm).^43,44 The AQE of SCS5 reached a maximum value of 16.80% at 475 nm, achieving the high efficiency with noble metal-free photocatalysts (Fig. 2f). Furthermore, the recycling test for SCS5 showed a small loss of photoactivity after continuous illumination lasting for 36 h as shown in Fig. 2g. This stable H₂ evolution for such a long illumination time suggests that CdS is stable and resistant to photocorrosion after coupling with the Co₃S₄/Co@C cocatalyst. The characterization results for SCS5 after the cycling tests are shown in Fig. S3 and (ESI).† No significant structural changes to SCS5 were observed in the SEM and TEM images after a long illumination. All the (101) facets of CdS, (111) facets of Co and (311) facets of Co₃S₄ were easily found. However, the peak intensities of CdS in the XRD patterns and XPS spectra decreased after the cycling tests. This may be caused by the slight photocorrosion of the CdS, leading to a small loss of photoactivity.

The UV-vis diffuse reflectance spectra (Fig. 2h) verified that the CdS exhibited a light absorption edge at about 550 nm. The Co₃S₄/Co-CdS photocatalysts showed the obvious improvement of visible light absorption, because of the black cocatalysts. Tauc plots (Fig. S5, ESI†) indicated that the Co₃S₄/Co@C cocatalysts had almost no influence on the band gap of the CdS. The positive tangential slopes in the Mott–Schottky plots confirmed the n-type semiconductor characteristics, as shown in Fig. S6 (ESI).† In addition, the smaller slopes in the Mott–Schottky plots of SCSx compared with those of CdS suggested a higher carrier density of SCSx due to the presence of the Co₃S₄/Co@C cocatalyst.⁴⁵ The smallest slopes of SCS5 mean the highest carrier density, which favors the photoelectrochemical and photocatalytic performance.⁴⁶ The EIS was performed under the visible-light illumination (>420 nm), as shown in Fig. 2i. Among all the samples, the SCS5 had the smallest hemicycle in the EIS Nyquist plots indicating the highest charge mobility. The EIS results confirmed that the Co₃S₄/Co-CdS photocatalysts were more favorable for separation and migration of the photogenerated charge carriers due to the excellent electrical conductivity of the Co₃S₄/Co@C cocatalyst, which was responsible for the superior photocatalytic performance of SCS5.

To investigate the origin of high photocatalytic activity on SCS5, the different Co@C and Co₃S₄/Co@C samples were inspected comprehensively. The XRD pattern of Co@C shows the typical diffraction peaks of metallic Co phases (Fig. 3a). After Co@C was vulcanized by TAA, the Co₃S₄ was successfully obtained with characteristic peaks at 2θ = 31.5° (311), 38.9° (400) and 55.1° (440). The TEM and HR-TEM analyses revealed that the metallic Co nanoparticles of Co₃S₄/Co@C retained their original structure when compared with that of Co@C (Fig. 3b), as shown in Fig. 3c. Furthermore, Co₃S₄ with the lattice spacings of 0.28 nm and 0.27 nm ascribed to (311) and (222) planes was adjacent to the Co nanoparticle, which indicated the intimate interface between Co and Co₃S₄ (Fig. 3d–e). These results demonstrated the coexistence of metallic Co and Co₃S₄ in Co₃S₄/Co@C, which was in agreement with the results shown in Fig. 1. The S 2p XPS peaks of the metal sulfide further confirmed the existence of Co₃S₄ in the Co₃S₄/Co@C cocatalyst (Fig. S7, ESI†), and the peaks at 168.5 eV and 169.6 eV were caused by the SO₄²⁻ species from the excess TAA.⁴⁷

Fig. 3 (a) XRD patterns. (b) TEM (inset) and HR-TEM images of Co@C. (c) TEM and (d and e) HR-TEM images of Co₃S₄/Co@C. (f) Co 2p XPS spectra of Co@C and Co₃S₄/Co@C.

Most importantly, the XPS spectra shown in Fig. 3f directly demonstrate the electron redistribution at the Co₃S₄/Co interface. The formation of Co–S covalent bonds will decrease the electron density of Co species in the Co₃S₄/Co@C sample, accompanying the shift to higher binding energies for the Co 2p XPS peaks.^32,48 This result proved the strong contact at the Co₃S₄/Co interface, which was beneficial for charge migration. In addition, compared to the C 1s XPS spectra of the Co@C and Co₃S₄/Co@C samples, there were no apparent changes for the graphitic-carbon structure,⁴⁹ as shown in Fig. S8 (ESI).†

As shown in Fig. 4a, the EIS strongly correlate with the charge transfer resistance on the Co@C and Co₃S₄/Co@C cocatalysts. The semicircular diameter in the Nyquist plot of Co₃S₄/Co@C was much smaller than that of Co@C, and the HER activity of the Co₃S₄/Co@C was much higher than that of Co@C (inset in Fig. 4a).⁵⁰ These results reveal that the Co₃S₄/Co@C cocatalyst has the faster charge transfer and kinetics for HER,⁵¹ which will boost the charge separation and migration in the photocatalyst. In general, the noble metal Pt is considered as a high efficiency cocatalyst for photocatalytic HER. However, the optimal sample of Pt-CdS (Fig. S9, ESI†) exhibited a lower photocatalytic H₂ evolution rate (5.40 mmol h⁻¹ g⁻¹) than that of SCS5 under the same conditions. The designed Co₃S₄/Co-CdS photocatalyst exhibited 2.8 times higher activity than Pt-CdS. Thus, it is believed that the Co₃S₄/Co@C cocatalyst is an ideal substitute for Pt.

Fig. 4 (a) Polarization curves for the HER (inset) and electrochemical impedance spectra of Co@C and Co₃S₄/Co@C in the dark. (b) Photocatalytic H₂ evolution, error bars represent the standard deviation of three tests. (c) Transient photocurrent responses. (d) PL spectra and electrochemical impedance spectra (inset) of CdS, C-CdS, Co-CdS, Co₃S₄-CdS and SCS5. (e) Surface photovoltage spectra, and (f) time-resolved PL spectra of CdS, Co-CdS and SCS5.

Next, different control samples were synthesized to prove the excellent photocatalytic H₂ evolution performance of Co₃S₄/Co-CdS, including the C-CdS (Fig. S10, ESI†), Co-CdS, and Co₃S₄-CdS (Fig. S11, ESI†) photocatalysts. Similarly, no intrinsic XRD peaks of Co and graphitic carbon were found in C-CdS, Co-CdS, and Co₃S₄-CdS (Fig. S12, ESI†), which were caused by the high-dispersion of Co₃S₄ and Co. The SCS5 sample also exhibited a stronger visible light absorption than that of C-CdS, Co-CdS, and Co₃S₄-CdS, as shown in Fig. S13 (ESI).† The photocatalytic H₂ evolution rate of SCS5 was much higher than those of C-CdS, Co-CdS and Co₃S₄-CdS (Fig. 4b), which demonstrated that the high performance of Co₃S₄/Co@C cocatalyst was due to the synergistic effect of Co and Co₃S₄. Furthermore, the (Co₃S₄/Co)-CdS sample was also synthesized by mechanically mixing the Co₃S₄/Co@C cocatalyst and CdS. The TEM analyses and EDX element mapping shown in Fig. S14 (ESI)† proved the coexistence of Co₃S₄/Co and CdS. The mechanically mixed (Co₃S₄/Co)-CdS also showed a much higher photocatalytic activity than that of Co-CdS and Co₃S₄-CdS, because of the effect of Co₃S₄. However, SCS5 showed a higher photocatalytic activity (Fig. S15, ESI†) than that of (Co₃S₄/Co)-CdS, which was probably caused by the larger contact interface of the cocatalyst and the semiconductor.

The efficient interfacial charge separation and migration between Co₃S₄/Co and CdS were confirmed by the transient photocurrent responses, PL spectra, EIS, SPV and time-resolved PL spectra. The SCS5 exhibited a much higher photocurrent density than those of Co₃S₄-CdS, Co-CdS, C-CdS and CdS (Fig. 4c), indicating the effective separation and migration of photogenerated electron–hole pairs in the Co₃S₄/Co-CdS structure. The lowest emission intensity in the PL spectra (Fig. 4d) further confirmed the highest electron–hole separation efficiency on CdS was obtained using Co₃S₄/Co@C as cocatalyst. The EIS in the inset of Fig. 4d shows that SCS5, with the smallest hemicycle in the Nyquist plot, was more resistant to charge recombination, which indicated a fast charge transfer between the Co₃S₄/Co@C cocatalysts and CdS under illumination. All these characterizations indicated that SCS5 has the fastest rate of interfacial charge separation and migration, resulting in the highest photocatalytic H₂ production rate. Furthermore, SPV was applied to investigate the photogenerated charge separation mechanism of the photocatalyst. The signal of the SPV spectrum depended on the type of semiconductor and direction of the migration of the photogenerated charge carriers.⁵² The positive SPV signal suggested a typical n-type semiconductor of CdS (Fig. 4e). The negative signal implied that photogenerated electrons were separated towards the cocatalysts in Co₃S₄-CdS, Co-CdS and SCS5.⁵³ The weakest SPV signal on SCS5 revealed that the Co₃S₄/Co@C cocatalyst served as a most efficient electron acceptor, which would offset part of the positive photoelectric SPV signal.⁵⁴ Meanwhile, the time-resolved PL spectra were used to obtain the charge carrier lifetimes of CdS coupling with different cocatalysts. As shown in Fig. 4f, the lifetime of the charge carrier on SCS5 (τ_av = 0.88 ns) was much shorter than those of Co-CdS (τ_av = 1.02 ns) and CdS (τ_av = 1.55 ns), signifying the effective charge migration from CdS to the Co₃S₄/Co@C cocatalyst.^55,56 Photocatalyst coupling with a cocatalyst was supposed to have a shorter lifetime owing to the fast charge transfer from photocatalyst to cocatalyst and suppression of the electron–hole recombination. The decrease of a lifetime for the SCS5 sample revealed that the bicomponent Co₃S₄/Co was a more effective cocatalyst for charge transfer compared with the one-component Co cocatalyst. The previous results confirmed the efficient interfacial charge separation and transfer in the Co₃S₄/Co-CdS photocatalyst.

The work functions of the cocatalyst were confirmed by the DFT calculations, as shown in Fig. 5a and b. The work functions of Co and Co₃S₄ were 4.93 eV and 5.87 eV, respectively. The larger work function signifies the more positive Fermi level (vs. vacuum level).⁵⁷ Hence, Co₃S₄ has a more positive Fermi level than that of Co, indicating that the photogenerated electrons tended to transfer to the Co₃S₄ from Co due to the different Fermi energy levels. The flat band energies from the Mott–Schottky plots vs. Ag/AgCl for CdS, Co-CdS and SCS5 were −1.04, −0.99 and −0.94 eV, respectively. Therefore, the Fermi levels vs. NHE (E_Ag/AgCl was about 0.197 eV vs. NHE, and the compensation potential was about 25 meV at room temperature) of CdS, Co-CdS and SCS5 were calculated to be −0.82, −0.77 and −0.72 eV, respectively. The more positive Fermi level of SCS5 was caused by the high working function of Co₃S₄. The conduction band (CB) position of CdS was generally more negative than the Fermi level by about −0.2 V,^37,58 therefore, the CB position of CdS was calculated to be −1.02 eV. Furthermore, these results were also consistent with the UPS spectra. The UPS spectra in Fig. S16a (ESI)† indicated that the work functions of Co@C and Co₃S₄@C were 3.72 and 4.31 eV (vs. vacuum level), respectively. Therefore, the Fermi levels vs. NHE (NHE was −4.44 eV vs. vacuum level)⁵⁹ of Co@C and Co₃S₄@C were calculated to be −0.72 and −0.13 eV, respectively, which further confirmed that the Co₃S₄ has a more positive Fermi level. Band gap structures of Co₃S₄/Co-CdS can be deduced from the previous analyses, as shown in Fig. S16b (ESI).† All the DFT calculations, Mott–Schottky plots and UPS proved that the most negative Fermi level of Co₃S₄ was in the Co₃S₄/Co-CdS photocatalyst, hence, the electrons were much easier to transfer to Co₃S₄, which had the active sites for photocatalytic hydrogen evolution.

Fig. 5 Fermi level calculated by DFT of (a) Co and (b) Co₃S₄. (c) Mott–Schottky plots of CdS, Co-CdS and SCS5. (d) Illustration of photogenerated carrier transfer between CdS and the Co₃S₄/Co@C cocatalyst.

Based on the previous analyses, the carriers transfer mechanism on Co₃S₄/Co-CdS for photocatalytic H₂ evolution is illustrated in Fig. 5d. Under the visible light irradiation, the photogenerated electrons were excited to the CB of CdS from the valence band (VB). Meanwhile, the holes in the VB were consumed by the sacrificial reagents on the surface of the CdS. As the conductive metal Co has a more negative Fermi level compared with the CB of CdS, the photogenerated electrons on CdS firstly transferred to the Co. In the Co₃S₄/Co@C cocatalyst, the Co₃S₄ has a more negative Fermi level than that of Co, leading to the transfer of electrons from Co to Co₃S₄. The electrons were captured by Co₃S₄ for the H₂ evolution. This stepwise energy band structure notably stimulates the charge separation and efficient electron transfer between the CdS and Co₃S₄/Co@C cocatalyst, due to the strong confinement effect of the electrons in Co₃S₄/Co. The Co₃S₄/Co@C cocatalyst acts as the electron sink, which was beneficial for the rapid migration of the photogenerated electrons and the high efficiency photocatalytic H₂ evolution reaction.

Conclusions

In summary, the bicomponent Co₃S₄/Co@C cocatalyst and CdS semiconductor were synthesized by a simple hydrothermal reaction using Co@C as a precursor. The novel Co₃S₄/Co-CdS heterostructure photocatalyst exhibited a high visible-light photocatalytic H₂ evolution activity of 15.17 mmol h⁻¹ g⁻¹, with a high quantum efficiency of 16.80% at 475 nm. The Co₃S₄/Co@C cocatalyst served as a high efficiency electron sink, accelerating the charge separation and migration between the interface of the Co₃S₄/Co@C cocatalyst and CdS. Furthermore, owing to the stepwise down energy band structure caused by the different Fermi levels of Co and Co₃S₄, the photogenerated electrons on CdS rapidly transfer to the Co₃S₄via Co metal, which can efficiently inhibit the recombination of electrons and holes in CdS. This work not only presents an inexpensive earth-abundant Co₃S₄/Co@C cocatalyst to replace noble metal cocatalysts, but also uncovers the huge potential of multicomponent cocatalysts in renewable solar energy production.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

We acknowledge the financial support from the National Natural Science Foundation of China (No. 21673080, 21802046, 51706231).

Notes and references

1. Y. Liu, Y. Li, S. Yang, Y. Lin, J. Zuo, H. Liang and F. Peng, ChemSusChem, 2018, 11, 2766–2775.
2. X. Chen, S. Shen, L. Guo and S. S. Mao, Chem. Rev., 2010, 110, 6503–6570.
3. Q. Zhao, J. Sun, S. Li, C. Huang, W. Yao, W. Chen, T. Zeng, Q. Wu and Q. Xu, ACS Catal., 2018, 8, 11863–11874.
4. Y. Miseki and K. Sayama, Adv. Energy Mater., 2018, 9, 1801294.
5. J. Ran, J. Zhang, J. Yu, M. Jaroniec and S. Z. Qiao, Chem. Soc. Rev., 2014, 43, 7787–7812.
6. G. Zhang, Z. A. Lan, L. Lin, S. Lin and X. Wang, Chem. Sci., 2016, 7, 3062–3066.
7. Y. Zhang, B. Tang, Z. Wu, H. Shi, Y. Zhang and G. Zhao, Green Chem., 2016, 18, 2424–2434.
8. A. T. Garcia-Esparza, D. Cha, Y. Ou, J. Kubota, K. Domen and K. Takanabe, ChemSusChem, 2013, 6, 168–181.
9. C. Guo, P. He, R. Cui, Q. Shen, N. Yang and G. Zhao, Adv. Energy Mater., 2019, 9, 1900364.
10. J. Ran, M. Jaroniec and S. Z. Qiao, Adv. Mater., 2018, 30, 1704649.
11. Y. Qu and X. Duan, Chem. Soc. Rev., 2013, 42, 2568–2580.
12. M. Tabata, K. Maeda, T. Ishihara, T. Minegishi, T. Takata and K. Domen, J. Phys. Chem. C, 2010, 114, 11215–11220.
13. R. Su, R. Tiruvalam, A. J. Logsdail, Q. He, C. A. Downing, M. T. Jensen, N. Dimitratos, L. Kesavan, P. P. Wells, R. Bechstein, H. H. Jensen, S. Wendt, C. R. Catlow, C. J. Kiely, G. J. Hutchings and F. Besenbacher, ACS Nano, 2014, 8, 3490–3497.
14. M. Hojamberdiev, Y. Cai, J. J. M. Vequizo, M. M. Khan, R. Vargas, K. Yubuta, A. Yamakata, K. Teshima and M. Hasegawa, Green Chem., 2018, 20, 3845–3856.
15. J. Yao, Y. Zheng, X. Jia, L. Duan, Q. Wu, C. Huang, W. An, Q. Xu and W. Yao, ACS Appl. Mater. Interfaces, 2019, 11, 25844–25853.
16. F. Zhang, H.-Q. Zhuang, J. Song, Y.-L. Men, Y.-X. Pan and S.-H. Yu, Appl. Catal., B, 2018, 226, 103–110.
17. G. Zhang, C. Huang and X. Wang, Small, 2015, 11, 1215–1221.
18. Z. Liu, Y. Huang, Q. Xiao and H. Zhu, Green Chem., 2016, 18, 817–825.
19. X. Zong, H. Yan, G. Wu, G. Ma, F. Wen, L. Wang and C. Li, J. Am. Chem. Soc., 2008, 130, 7176–7177.
20. Y. Zheng, J. Dong, C. Huang, L. Xia, Q. Wu, Q. Xu and W. Yao, Appl. Catal., B, 2020, 260, 118220.
21. J. Dong, Y. Shi, C. Huang, Q. Wu, T. Zeng and W. Yao, Appl. Catal., B, 2019, 243, 27–35.
22. J. D. Xiao, L. Han, J. Luo, S. H. Yu and H. L. Jiang, Angew. Chem., Int. Ed., 2018, 57, 1103–1107.
23. X. She, H. Xu, L. Li, Z. Mo, X. Zhu, Y. Yu, Y. Song, J. Wu, J. Qian, S. Yuan and H. Li, Appl. Catal., B, 2019, 245, 477–485.
24. Y.-F. Xu, M.-Z. Yang, H.-Y. Chen, J.-F. Liao, X.-D. Wang and D.-B. Kuang, ACS Appl. Energy Mater., 2018, 1, 5083–5089.
25. S. Bai, J. Jiang, Q. Zhang and Y. Xiong, Chem. Soc. Rev., 2015, 44, 2893–2939.
26. H. Husin, W.-N. Su, H.-M. Chen, C.-J. Pan, S.-H. Chang, J. Rick, W.-T. Chuang, H.-S. Sheu and B.-J. Hwang, Green Chem., 2011, 13, 1745–1754.
27. Y. Ma, R. Chong, F. Zhang, Q. Xu, S. Shen, H. Han and C. Li, Phys. Chem. Chem. Phys., 2014, 16, 17734–17742.
28. Q. Fu and X. Bao, Chem. Soc. Rev., 2017, 46, 1842–1874.
29. A. Meng, L. Zhang, B. Cheng and J. Yu, Adv. Mater., 2019, 31, 1807660.
30. M. Bowker, D. James, P. Stone, R. Bennett, N. Perkins, L. Millard, J. Greaves and A. Dickinson, J. Catal., 2003, 217, 427–433.
31. S. Bai, L. Yang, C. Wang, Y. Lin, J. Lu, J. Jiang and Y. Xiong, Angew Chem., Int. Ed., 2015, 54, 14810–14814.
32. Z.-H. Xue, H. Su, Q.-Y. Yu, B. Zhang, H.-H. Wang, X.-H. Li and J.-S. Chen, Adv. Energy Mater., 2017, 7, 1602355.
33. X. Dai, M. Xie, S. Meng, X. Fu and S. Chen, Appl. Catal., B, 2014, 158–159, 382–390.
34. K. Zhang, J. Ran, B. Zhu, H. Ju, J. Yu, L. Song and S. Z. Qiao, Small, 2018, 14, e1801705.
35. S. Grimme, J. Comput. Chem., 2006, 27, 1787–1799.
36. A. Klamt and G. Schüürmann, J. Chem. Soc., Perkin Trans. 2, 1993, 799–805.
37. H. Guan, S. Zhang, X. Cai, Q. Gao, X. Yu, X. Zhou, F. Peng, Y. Fang and S. Yang, J. Mater. Chem. A, 2019, 7, 2560–2574.
38. B. Chen, R. Li, G. Ma, X. Gou, Y. Zhu and Y. Xia, Nanoscale, 2015, 7, 20674–20684.
39. G. Zhao, W. Zhou, Y. Sun, X. Wang, H. Liu, X. Meng, K. Chang and J. Ye, Appl. Catal., B, 2018, 226, 252–257.
40. W. Gu, L. Hu, W. Hong, X. Jia, J. Li and E. Wang, Chem. Sci., 2016, 7, 4167–4173.
41. J. Du, T. Zhang, J. Xing and C. Xu, J. Mater. Chem. A, 2017, 5, 9210–9216.
42. L. Liu, A. V. Puga, J. Cored, P. Concepción, V. Pérez-Dieste, H. García and A. Corma, Appl. Catal., B, 2018, 235, 186–196.
43. S. Chen, Y. Qi, T. Hisatomi, Q. Ding, T. Asai, Z. Li, S. S. Ma, F. Zhang, K. Domen and C. Li, Angew. Chem., Int. Ed., 2015, 54, 8498–8501.
44. G. Zhao, Y. Sun, W. Zhou, X. Wang, K. Chang, G. Liu, H. Liu, T. Kako and J. Ye, Adv. Mater., 2017, 29, 1703258.
45. Y. Liu, Y. Li, F. Peng, Y. Lin, S. Yang, S. Zhang, H. Wang, Y. Cao and H. Yu, Appl. Catal., B, 2019, 241, 236–245.
46. K. Zhang, W. Zhou, L. Chi, X. Zhang, W. Hu, B. Jiang, K. Pan, G. Tian and Z. Jiang, ChemSusChem, 2016, 9, 2841–2848.
47. H. Wang and Z. Jin, Sustainable Energy Fuels, 2019, 3, 173–183.
48. S. Liu, X. Zhang, G. Wang, Y. Zhang and H. Zhang, ACS Appl. Mater. Interfaces, 2017, 9, 34269–34278.
49. P. Chen, T.-Y. Xiao, Y.-H. Qian, S.-S. Li and S.-H. Yu, Adv. Mater., 2013, 25, 3192–3196.
50. X. Huang, Q. Shen, J. Liu, N. Yang and G. Zhao, Energy Environ. Sci., 2016, 9, 3161–3171.
51. X. Xu, F. Nosheen and X. Wang, Chem. Mater., 2016, 28, 6313–6320.
52. R. Chen, F. Fan, T. Dittrich and C. Li, Chem. Soc. Rev., 2018, 47, 8238–8262.
53. T. K. Townsend, N. D. Browning and F. E. Osterloh, Energy Environ. Sci., 2012, 5, 9543–9550.
54. X. Yue, S. Yi, R. Wang, Z. Zhang and S. Qiu, Small, 2017, 13, 1603301.
55. W. Tu, Y. Zhou, Q. Liu, S. Yan, S. Bao, X. Wang, M. Xiao and Z. Zou, Adv. Funct. Mater., 2013, 23, 1743–1749.
56. Z. Sun, H. Zheng, J. Li and P. Du, Energy Environ. Sci., 2015, 8, 2668–2676.
57. S. Cao, B. Shen, T. Tong, J. Fu and J. Yu, Adv. Funct. Mater., 2018, 28, 1800136.
58. S. Chandrasekaran, L. Yao, L. Deng, C. Bowen, Y. Zhang, S. Chen, Z. Lin, F. Peng and P. Zhang, Chem. Soc. Rev., 2019, 48, 4178–4280.
59. S. Trasatti, Pure Appl. Chem., 1986, 58, 955–966.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9gc03323h

---