Introduction

Semiconductor photocatalysis as a green technology for wastewater/organic contaminants treatment and green energy production has attracted considerable attention since Fujishima and Honda realized water splitting to generate hydrogen by using TiO2 in 19721,2,3,4. Since then, various photocatalysts, such as transition metal oxides5,6,7,8, metal sulfides9,10,11,12, heterojunctions13,14,15, doping materials16,17,18, composite structures19, 20, have been synthesized to improve the photocatalytic activities. Among them, metal chalcogenides were considered to be promising photocatalytic candidates due to their unique properties, such as suitable band gap, ideal electronic band position and thus exhibiting excellent catalytic activities10, 21, 22. Particularly, copper sulfide (CuS) and cadmium sulfide (CdS) were widely investigated in the application of photocatalysis4, 10, 23,24,25,26.

CuS could absorb visible light in the solar spectrum due to its narrow band gap of about 2.08 eV4, 10, 27, 28 and the abundant raw materials on earth have made the photocatalysts syntheses inexpensive. Therefore, considerable attention and approaches have been paid to synthesize diverse shaped CuS. For instance, CuS microflowers composed of nanosheets were obtained by a one-pot sonochemical process and their versatile photocatalyst responses were investigated under natural light irradiation by Cao et al.4. Pradhan et al. has synthesized CuS submicro-spheres and nanotubes by a solution chemistry route and revealed that the synthetic conditions could affect the shape, size and structure of CuS and thus its photocatalytic activities10. However, using CuS as photocatalyst alone was similar to most single semiconductor photocatalyst which encounters low photocatalytic activity3, 29. Therefore, many efforts have been made to improve the photocatalytic properties of CuS-based photocatalysts. For example, metal ions doped CuS with enhanced visible light photocatalytic activity on dyes degradation was prepared by Hosseinpour et al.28. In addition, CuS loaded on metal sulfides including ZnS30, 31, CdS29, even ZnS-CuS-CdS composite21, were studied and found possessing high photocatalytic activity toward H2 generation under visible light irradiation. So far, exploring more efficient CuS-based photocatalyst is still highly desirable because few studies based on CuS as the host coupled with metal sulfides are conducted.

Although the rapid recombination rate of photogenerated electron-hole pairs for pure CdS on one hand limits its photocatalytic activity owing to its band energies22, 29, 32,33,34,35,36, the serious photocorrosion of CdS is another obstacle hindering its wide application as high-performance photocatalysts34, 36. But, CdS is still worthwhile to be used as a photocatalyst due to its narrow band gap of 2.4 eV, endowing it the extremely feasible feature to absorb light irradiation in visible light range22, 24, 29. Therefore, many efforts have been made to overcome the aforementioned disadvantages and various nanostructured CdS materials had been synthesized, such as nanorods23, nanoflowers23, nanospheres24, nanotubes37, nanowires38, 39, nanosheets40, nanocone and nanofrustum41 etc. Recently, the photocatalytic efficiency of CdS could be further improved by coupling it with other materials to form hybrid structures, such as WS2 22, ZnO33, Al2O3 33, ZnS-CuS21, CuS29, TiO2 42, 43, graphene32, histidine44 etc, due to the effective separation of photogenerated electron-hole pairs. Therefore, it is reasonable to deduce that the structure hybridization of CuS and CdS could award us distinctly improved photocatalytic activities by extending light absorption of solar spectrum29, 33. In this sense, the facile preparation of high quality CdS loaded CuS heterostructure with precisely controlled morphologies and compositions and systematically investigate its photodegradation efficiency is very intriguing and important.

In this work, CdS decorated CuS photocatalysts were rationally designed and controllably synthesized via a facile one-pot hydrothermal method. The effect of the CdCl2 and the concentration of thiourea precursor on the structures and morphologies of as-prepared samples were systematically investigated by different characterization techniques. The photocatalytic activities of the as-prepared samples over methyl orange (MO) degradations under visible light irradiation were performed and a possible reaction mechanism was proposed by evaluating the post-photodegradation analysis.

Experimental

Synthesis of CdS decorated CuS photocatalysts

All chemical reagents were of analytical grade and used without further purification, which purchased from Sinopharm Chemical Reagent Co., Ltd. (SCRC, China). Typically, a certain amount of cadmium chloride (CdCl2 · 2.5H2O), 3 mmol thiourea and 0.098 g polyethylene glycol (PEG, Mw = 2000) were dissolved into 80 mL deionized water by vigorously magnetic stirring for 10 min. Then, 4 mmol copper (II) nitrate trihydrate (Cu(NO3)2 · 3H2O) was added into the solution followed by magnetically stirring for 1 hour to form homogeneous solution. After that, the solution was transferred into 100 mL Teflon-lined stainless steel autoclave. Thereafter, the sealed autoclave was kept at 140 °C for 10 hours, followed by cooling down to room temperature naturally. Subsequently, the as-prepared precipitants were collected by centrifugation and washed with deionized water and ethanol for several times. Finally, the products were obtained after drying the precipitants at 60 °C for 12 hours in a vacuum oven. The samples were named as Cd-0, Cd-0.5, Cd-1, Cd-2, and Cd-4 with the amount of CdCl2 · 2.5H2O of 0, 0.5, 1, 2, and 4 mmol, respectively. For comparison, another three samples (Cd-1-T4, Cd-1-T6, and Cd-0-T6) were prepared by changing the amount of chemical reagents while keeping other conditions the same. All the preparation conditions of the samples are listed in Table 1.

Characterization

X-ray powder diffraction (XRD) patterns of as-prepared samples were recorded by a German X-ray diffractiometer (D8-Advance, Bruker AXS, Inc., Madsion, WI, USA) equipped with Cu K α radiation (λ = 0.15406 nm). The morphologies of the samples were observed by a field emission scanning electron microscope (FESEM, FEI Quanta FEG250, FEI, Hillsboro, USA) and transmission electron microscope (TEM, HEOL-200CX, JEOL, Tokyo, Japan). High-resolution TEM images were also investigated with a field transmission electron microscope (Tecnai G2 F20 S-TWIN, FEI, Hillsboro, USA). The X-ray photoelectron spectroscopy (XPS) was collected on the Thermo ESCALAB 250XI electron spectrometer equipped with Al K α X-ray radiation ( = 1486.6 eV) as the source for excitation (ThermoFisher Scientific, Waltham, MA USA). The Brunauer–Emmett–Teller (BET) specific surface areas of as-prepared samples were measured by N2 adsorption–desorption isotherm with a Quantachrome NOVAtouch LX4 apparatus (Quantachrome Instruments, South San Francisco, CA, USA).

Photocatalytic measurement

The photocatalytic properties of as-prepared samples were characterized by a UV-vis spectrophotometer (TU-1901, Beijing Purkinje General Instrument Co., Ltd, Beijing, China) at room temperature in air under visible light irradiation, which was similar to previous reports8, 16. The visible light was generated by a 500 W Xe lamp equipped with a cutoff filter (λ ≥ 420 nm) to remove the UV part. A typical process was carried out as follows: 30 mg products were dispersed into 50 ml of 10 mg/L methyl orange (MO) aqueous solution. Then, the suspension was kept in dark for 30 min with magnetic stirring to reach adsorption-desorption equilibrium of MO on the surface of as-grown samples. After a given irradiation time interval, ca. 3 mL suspension was transferred into the centrifuge tube to separate the powders and solution for the purpose of UV-vis spectra test. The concentration of MO was evaluated by measuring the absorbance properties at 464 nm in UV-vis spectra, which was used to illuminate the photocatalytic properties of as-prepared samples.

Results and Discussion

Structural and morphological characterization

XRD patterns of the as-prepared samples in Figs 1 and 2 show that all the peaks can be perfectly indexed into hexagonal CuS phase (JCPDS No. 65–3556). No other characteristic peaks could be found for all the samples, demonstrating their high crystalline purity and lower-level loading content of CdS on CuS. The only differences observed from these patterns were the intensity increase of (100) peak and the intensity decrease of (101) and (006) peaks, which could be ascribed to the addition of Cl ions and the amount change of thiourea during the synthetic process that result in a tiny stoichiometry vary of the copper sulfides45. However, the phase and composition of the as-prepared samples keep almost unchanged with the addition of CdCl2 and thiourea during the preparation process.

Sample Cd-0 in Fig. 3a presents the hierarchical structures consisted of nanofibers, spheres, and flake-like morphology, illustrating the uncontrollable morphologies manner of the sample synthesized without the existence of CdCl2. Interestingly, after addition CdCl2 (sample Cd-0.5), the morphology transformed into flower-like structure with inhomogeneous sizes and thin petals that composed of compact nanosheets, as depicted in Fig. 3b. Further increasing the amount of CdCl2 (sample Cd-1) results in the formation of compact microflower-like structure with porous petals (Fig. 3c). Then, the compact flowers evolved into loosened structures accompanied by the decrease of porosities with further CdCl2 content increasing (sample Cd-2, Fig. 3d). Finally, the morphology transformation completed with a result of assembling the nonporous and thick sheets together to form the heterogeneous structure for sample Cd-4 (Fig. 3e). In order to investigate the element distribution of the as-prepared samples, the energy dispersive X-ray spectroscopy (EDS) elemental mappings of two typical products Cd-1 and Cd-4 were recorded (Fig. 4). The results show that S, Cu and Cd elements are homogeneously distributed throughout the flowers. Furthermore, the effect of thiourea used on the morphology of the as-prepared samples was investigated. As shown in Fig. 5, by fixing the CdCl2 content (1 mmol) during the synthetic process, the porosity was disappeared with the increment use of thiourea (Fig. 5a, sample Cd-1-T4) and the sheets forming the petals became thicker as seen from the magnified SEM image (inset of Fig. 5a). After further increasing the amount of thiourea (Cd-1-T6, Fig. 5b) the flower becomes more compact and the petals grows even thicker with the disappearance of porosity. For comparison, the sample using 6 mmol thiourea without CdCl2 addition (Cd-0-T6) was prepared and its SEM morphology was presented in Fig. 5c and d. It can be seen that fibers and irregular particles were formed which was quite different from that of Cd-0. Taken together, these results demonstrated that both thiourea and Cl anions from CdCl2 addition (see Supporting Information Fig. SI-1) played important roles in regulating the morphology of the as-prepared samples, in well agreement with previous reports45,46,47,48,49.

TEM observation was further used to investigate the details of the as-prepared products. Figure 6a and b show that microflower structures with porous thin petals for Cd-1 and thick petals with a lower porous density for sample Cd-2 in Fig. 6c and d are observed, respectively. Figure 6e and f clearly confirm the existence of CdS particles deposited on the surface of CuS, and the distance between two crystal lattice fringes are 0.322 nm (CuS (101)) and 0.336 nm (CdS (002)) consisting with previously reported result29. In addition, the particle density in sample Cd-2 is higher than that of Cd-1, indicating more CdS particles were formed. Moreover, the thickness of the petals for Cd-1-T4 get thicker and the porosity drastically decreases (Fig. 6g and h), while the porosity almost disappears and the petals developed more thickly for Cd-1-T6 as shown in Fig. 6i and j. All these TEM results are consistent with the aforementioned SEM observations.

The XPS spectra of the prepared samples were recorded further to confirm the composition and the elemental oxidation states, as depicted in Fig. 7. The survey scans show that all the main peaks could be indexed into Cu, S, and Cd elements for all samples though the intensities were different (see Supporting Information Fig. SI-2). Particularly, the Cd peaks could not be detected for samples Cd-0 and Cd-0-T6, in well agreement with the synthesis conditions, where the absence of Cd addition in the reaction solution. The two stronger peaks with an energy separation of 19.9 eV at around 932.0 eV and 951.9 eV from Cu 2p region in Fig. 7a and b, respectively, are in accordance with the binding energy peaks for Cu 2p3/2 and Cu 2p1/2, conforming the Cu oxidation state is Cu (II)4, 29. The weak shakeup satellite peaks at 943.0 eV and 962.8 eV are also indexed to Cu2+ ions, indicating the paramagnetic chemical state of Cu2+ ion4, 8, 29. The Cd 3d region (except for sample Cd-0-T6) in Fig. 7a and b could be fitted into two main peaks locating at 405 eV and 412 eV, which are assigned to the binding energies of Cd 3d5/2 and Cd 3d3/2, suggesting the existence of Cd2+ in CdS22, 29, 50. In terms of Cd-0-T6 in Fig. 6b, there is no peak observed in the high resolution Cd 3d XPS region, illustrating the non-existence of CdS, which agrees well with the preparation conditions. The core-level XPS spectra of S 2p show different behaviors among the samples as depicted in Fig. 7. For the samples without CdCl2 addition (Cd-0 and Cd-0-T6), only two main peaks located at 162.3 eV (S 2p3/2) and 163.5 eV (S 2p1/2) are observed which are assigned to S2−, indicating the presence of metal sulfides (CuS)4. Another peak at around 161.5 eV appeared after the addition of CdCl2 for the XPS spectra of S 2p region, confirming the formation of metal sulfides including Cu2S29, 51, 52. Moreover, the peak positions of Cu 2p1/2 at 951.9 eV and S 2p at 161.5 eV are slightly shifted to lower binding energy region with the increase of CdCl2 content, indicating the existence of more Cu+ on the surface in Cd-0.5, Cd-1, Cd-2, and Cd-4. In addition, the peak positions of Cu 2p for Cd-1-T4 are apparently shifted to much lower binding energy region, compared with those of Cd-0-T6 and Cd-1-T6, illustrating the formation of much more Cu+ on the surface. These results are in well agreement with the observation of S 2p region and previous reports30, 52. The reason for the formation of Cu2S may be ascribed to the amount of thiourea and chloride ions from the addition of cadmium chloride in the precursor solution45, 52,53,54.

The growth mechanism for the formation CdS decorated CuS could be explained based on previous reports4, 22, 29, 45, 50,51,52,53,54. Thiourea (Tu) is considered to easily coordinate with copper (II) ion in aqueous solution owing to the availability of lone pair of electrons on the ligand and free states of metallic ions4, 22, 55. Then thiourea-copper (II) complex could serve as Cu2+ precursors. The possible chemical reaction could be expressed as follows4, 10, 55,56,57:

$${{\rm{Cu}}}^{2+}+x{\rm{Tu}}\to {[{\rm{Cu}}{({\rm{Tu}})}_{x}]}^{2+}$$
(1)
$${{\rm{NH}}}_{2}{{\rm{CSNH}}}_{2}+2{{\rm{H}}}_{2}{\rm{O}}\to {{\rm{CO}}}_{2}+2{{\rm{NH}}}_{4}^{+}+{S}^{2-}$$
(2)
$${{\rm{Cu}}}^{2+}+{S}^{2-}\to {\rm{CuS}}$$
(3)
$${{\rm{Cd}}}^{2+}+{S}^{2-}\to {\rm{CdS}}$$
(4)

The different products solubility of CdS (8 × 10−28) and CuS (6.3 × 10−36) would generate the preferential formation of CuS followed by CdS, which generate the CdS decorated CuS structure51. The appearance of Cu2S after the addition of cadmium chloride (CdCl2) mainly caused by the introduction of chloride anion (Cl) with the assist of polyethylene glycol under hydrothermal condition, which have been reported by other groups45, 53, 54, 58. The higher concentration of chloride ions in the precursor solution facilitates the formation of Cu2S, interpreting the aforementioned tendency of binding energy shift. When more thiourea was added into the precursor solution, the amount of Cu2S phase decreases, which could be explained using the following reaction53, 59:

$${{\rm{Cu}}}_{2}{\rm{S}}+{S}^{2-}\to 2{\rm{CuS}}+2{e}^{-}$$
(5)

Photocatalytic properties

$$\mathrm{ln}(C/{C}_{0})=-kt$$