1 Introduction

The mitigation energy crisis has attracted numerous research interests in the field of green energy and renewable sources in recent years [1,2,3]. Photocatalytic hydrogen generation can be regarded as a promising way to solve this dilemma due to its renewability, stabilization and cleanliness. In particular, the hollow nanostructured materials are viewing a novel model because it can enhance the light absorption, reagent enrichment and efficient interfacial reaction during the photoinduced hydrogen evolution reaction (HER). Therefore, the nano-sized hollow structure and its derivative nanomaterials such as core–shell, yolk–shell and ball-in-ball-structured semiconductor materials have been applied in energy storage, catalysis and electrochemistry [4,5,6]. Among the above multilevel structure, the classical core–shell hollow structure has the follow advantages. Firstly, when the light was penetrated into the nano-sized hollow structure, it would trigger the multiple scattering effect because the photon could be scattered by the inner shell and the pore structure for the enhanced utilization of proton energy compared with other morphology of photocatalysts. Secondly, the hollow structure would provide a relative homogeneous reaction space for boosting the HER activity. Finally, the hollow structure could offer the larger specific surface area and pore structure could increase the mass transfer efficiency or active sites during the H2 evolution [4]. Hence, it is highly expected to establish a novel core–shell hollow nanostructured material for effective photocatalytic H2 generation.

Titanium dioxide (TiO2) with the formidable defect of low light utilization and relative wide band gap (3.20 eV) greatly limited its practical application. Hence, a plenty of efforts are paid to improve its performance, including the lattice-doping, morphological control and heterogeneous structure construction. In term of the above research, Wang et al. [7] have reported the growth of ZnFe2O4 nanoparticles on TiO2 nanotube arrays for enhanced visible-light photocatalytic activity. Xiang et al. [8] have prepared the graphene-modified titania nanosheets for enhanced photocatalytic H2 production activity. Yang et al. [9] have reported the novel three-dimensionally ordered macroporous Fe3+-doped TiO2 photocatalysts for H2 production and degradation applications. Nevertheless, due to the relatively low charge transfer efficiency, the above photocatalysts still cannot satisfy the practical application requirement. Therefore, it is desired to synthesize the hollow-structured TiO2 via anchoring with ternary chalcogenide to form heterojunction and thus enhance the charge separation and migration under overall light irradiation. ZnIn2S4 as a typical photocatalyst has been widely concerned due to its great optical, electrical properties, suitable band gap (1.99 eV) and promising stability [10]. Therefore, the construction of TiO2@ZnIn2S4 core–shell hollow nanospheres could highly improve the transfer efficiency of charge carriers in the heterojunction system. Simultaneously, multiple scattering effect depended on hollow nanostructure could boost the light absorption of catalyst, as a result of leading to high HER activity.

Herein, according to the above theoretical analysis, in this work, we report the formation of TiO2@ZnIn2S4 hollow nanospheres which exhibit a great photocatalytic hydrogen evolution under overall light irradiation. This work is expected to pave a way to synthesize the novel hollow core–shell structure, and the detailed mechanism is proposed to explain the migration of photoinduced electrons and holes in this system.

2 Experimental

2.1 Chemical and materials

All chemical reagents were of analytical grade and used without any further purification, including tetraethyl orthosilicate (TEOS, ≥ 99.5%), tetrabutyl titanate (TBOT, ≥ 98%), potassium chloride (KCl, ≥ 99.5%), indium nitrate (In(NO3)3, ≥ 99.99%), zinc nitrate hexahydrate (Zn(NO3)2·6H2O, 99.998%), thioacetamide (TAA, 98%), 2-propanol (C3H8O, ≥ 99.5%), methyl alcohol (CH3OH, 99%), ammonium hydroxide (NH3·H2O, ≥ 99.5%), sodium hydroxide (NaOH, 99%).

2.2 Synthesis of SiO2 nanospheres

The monodisperse SiO2 template was synthesized by the Stöber method. Briefly, 40 ml 2-propanol, 1.5 ml ammonia solution and 2.0 ml deionized water were mixed and stirred vigorously in a flask. And then, 3.5 ml TEOS was added rapidly and stirred for 24 h at room temperature. Afterward, the predecessor was collected by the centrifugation and washed several times with deionized water and dried in a 60 °C oven for 5 h.

2.3 Preparation of TiO2 hollow nanospheres

The synthesis of TiO2 hollow nanospheres was prepared by the following method. First, 40 ml methyl alcohol and 0.3 ml potassium chloride solution (0.1 mol·L−1) were ultrasonically dissolved and added 0.1 g SiO2 nanospheres in a reaction vessel. Then, 0.5 ml TBOT was added rapidly and stirred for 4 h at room temperature. The samples were washed several times with ethyl alcohol and dried in a 60 °C oven for 5 h and then calcined at 500 °C for 1 h in a muffle furnace. After the products were cooled down to room temperature, NaOH aqueous solution (10 mol·L−1) was added into the solution to etch SiO2 template and the products were washed with deionized water and dried in a 60 °C oven for 5 h.

2.4 Preparation of TiO2@ZnIn2S4 hollow nanospheres

The TiO2@ZnIn2S4 hollow nanospheres were prepared by the hydrothermal method. The optimal procedures were described as follows: 0.5 mmol cadmium chloride, 1.0 mmol thiourea and 4.0 mmol thioacetamide were dissolved in a flask containing 40 ml deionized water. Afterward, the certain amount of TiO2 hollow nanospheres was added into the above mixture and stirred for 4 h. Then, the solution was transferred to a 100-ml Teflon-lined stainless steel autoclave which was heated to 180 °C for 3 h, it was then cooled to room temperature and washed with deionized water and ethanol several times. Afterward, the samples were dried in a 60 °C oven for 5 h.

2.5 Preparation of TiO2/ZnIn2S4 composite

The synthesis of TiO2 cluster was carried out by the following method. Typically, 30.0 ml methyl alcohol and 0.5 ml deionized water were mixed in a reaction vessel. Then, 1.0 ml TBOT was added rapidly and vigorously stirred for 4 h at room temperature. Afterward, the predecessor was washed several times with deionized water, dried in a 60 °C oven for 4 h and then calcined at 500 °C for 1 h in a muffle furnace. After the products were cooled down to room temperature, 0.5 mmol cadmium chloride, 1.0 mmol thiourea and 4.0 mmol thioacetamide were dissolved in a flask containing 40 ml deionized water stirring for 4 h. Then, the following process was the same with the preparation of TiO2/ZnIn2S4 nanospheres.

2.6 Characterization

The experimental samples were tested by various measurements. The crystal phase of as-prepared samples was analyzed using X-ray diffractometer (XRD, Lab X XRD-6100, SHIMADZU). The structure and morphology of the products were investigated using a field emission scanning electron microscope (FESEM, JSM-6700F, JEOL) equipped with an energy-dispersive spectrometer (EDS) and transmission electron microscope (TEM, JEM-2100, JEOL). The ultraviolet–visible (UV–Vis) spectra over the samples were obtained by a UV–Vis spectrophotometer (UV-2600, SHIMASZU). The electrochemical measurements were obtained to test on a CHI660E electrochemical apparatus (Shanghai, China) using a standard three-electrode cell equipped a working electrode, a platinum wire and Ag/AgCl as the counter electrode and reference electrode, respectively. The electrolyte was 1.0 mol·L−1 Na2SO4 solution. The surface photovoltaic spectrum (SPS) was carried out using a solid junction cell with a light source-monochromator-lock-in detection apparatus. Then, the samples were irradiated by the 500-W xenon lamp with a triple-prism monochromator (Hilger and Watt, D300). And an amplifier (Brookdeal, 9503-SC) was synchronized with a light chopper, which was obtained to amplify the signal of photovoltage. The SPS signals were normalized to unity at the characteristic peaks, which were measured by a laptop, respectively. All the SPS measurements were performed at room temperature.

2.7 Photocatalytic H2 evolution

The photocatalytic water splitting reaction was analyzed in a Pyrex reactor for photoinduced irradiation at room temperature. Typically, 0.03 g sample was dispersed in 40 ml aqueous solutions that contain 5 ml lactic acid (sacrificial electron donor). And the reaction vessel was deaerated by the bubble treatment using the high purity nitrogen. Then, the evolved gases were evaluated by an online gas chromatography (SP-2100A, Beifen-Ruili) with the thermal conductivity detector (TCD) detector and the photocatalytic reaction was tested under the irradiation of a 300-W xenon lamp (HSX-F300, Nbet Beijing).

The recycled experiments of photocatalytic reactions were analyzed under the above conditions to test the sustainable property. After 4-h photocatalytic H2 evolution reaction, the suspensions were washed and centrifuged by ethanol several times and then dried in an oven for 3 h at 50 °C. The drying photocatalyst was re-dispersed in 40 ml deionized water with 5 ml lactic acid in reaction vessel, respectively. Before the photoreaction, the system was degassed to remove the air in the reaction solution.

The apparent quantum yield (AQY) of photocatalytic H2 production was tested on the similar condition. A xenon lamp with different monochromatic lights was as the light source (λ = 435 nm). The illuminated area on the reaction cell was about 4.52 cm2. The average photocatalytic H2 generation rate was measured after 3 h. Afterward, the AQY for H2 evolution was calculated according to the following equation [4]:

$${\text{AQY = }}\frac{{N_{\text{re}} }}{{N_{\text{ip}} }} \times 100\% = \frac{{2N_{\text{ehm}} }}{{N_{\text{ip}} }} \times 100\%$$
(1)

where Nre is the number of reactor electrons, Nip is the number of incident photons and Nehm is the number of evolved hydrogen molecules.

3 Result and discussion

The overall synthetic process of TiO2@ZnIn2S4 hollow nanospheres is shown in Fig. 1. Firstly, monodisperse SiO2 nanospheres were prepared by a previously reported Stöber method with modification [4]. And then, the solvent–thermal method was adopted to synthesize SiO2@TiO2 core–shell composite. Afterward, the resulted white powder underwent ultrasonic exfoliation to obtain homogeneous solution, while the selective etching of SiO2 inner core was used to obtain the TiO2 hollow nanospheres. Then, the precursors of metal ions Zn2+ and In3+ can be preloaded onto the surface of TiO2 hollow nanospheres owing to their electrostatic interaction. Next, under the hydrothermal reduction process, the thioacetamide acted as a sulfur source would react with Zn2+ and In3+ to in situ generate and anchor ZnIn2S4 nanoleaves on the surface of TiO2 nanospheres, which finally leads to the formation of TiO2@ZnIn2S4 hollow heterojunction.

Fig. 1
figure 1

Schematic representative of formation procedure for as-synthesized TiO2@ZnIn2S4 hollow nanospheres

Full size image

The crystal phase of the as-synthesized composites was measured by XRD. As shown in Fig. 2a, the main characteristic peaks at 25.2°, 38.5°, 48.0°, 55.0° and 62.6° in the curve of the pure TiO2 can be, respectively, assigned to the (101), (004), (200), (211) and (204) diffraction plane of the anatase structure TiO2 (JCPDS No. 21-1272). And the three main characteristic peaks located at 21.5°, 28.3° and 54.8° correspond to the (009), (104) and (021) diffraction plane of the rhombohedral structure ZnIn2S4 (JCPDS No. 49-1562), respectively [11, 12]. Specially, Fig. 2b shows the typical high-resolution TEM (HRTEM) of the TiO2@ZnIn2S4 hollow nanospheres, which can be matched to the lattice fringes of 0.342 and 0.411 nm, corresponding to the (101) crystal plane of anatase TiO2 and the (009) crystal plane of rhombohedral ZnIn2S4, respectively. In addition, the corresponding EDS data are displayed in Fig. 2c, d, and the titanium (Ti), oxygen (O), zinc (Zn), indium (In) and sulfur (S) elements can be observed. These above results demonstrate that ZnIn2S4 nanoleaves are uniformly anchored on the surface of TiO2 hollow nanospheres and strongly support that the novel TiO2@ZnIn2S4 hollow nanospheres are successfully obtained.

Fig. 2
figure 2

XRD patterns of a as-synthesized samples and b HRTEM image of optimal sample (TiO2@ZnIn2S4 hollow nanospheres); c TEM image and d EDS result of optimal sample

Full size image

Figure 3a–d show the SEM images of the as-prepared samples. It can be seen that the SiO2 template exhibits the typically spherical morphology with the average diameter of 450 nm in Fig. 3a. And the pure TiO2 hollow nanosphere has an obvious inner cavity with the average of diameter of 720 nm in Fig. 3b. The pure ZnIn2S4 exhibits the micron-sized flower-like structure with the average diameter of 4 μm. Figure 3d shows the SEM image of the TiO2/ZnIn2S4 cluster sample. As shown, a typical irregular TiO2 agglomerates and the flower-like ZnIn2S4 nanoleaves are uniformly dispersed on the surface of TiO2 cluster. Figure 3e, f clearly show the TEM images of TiO2@ZnIn2S4 hollow nanospheres. The as-obtained sample possesses typical spatial cavity with the average diameter of 520 nm and outer shell with about 100 nm in thickness, which is beneficial for the light absorption and reactant enrichment in their inner cavity, leading to the great enhancement of the photocatalytic H2 generation [13].

Fig. 3
figure 3

SEM images of a SiO2 nanospheres b TiO2 hollow nanospheres, c pure ZnIn2S4 cluster and d TiO2/ZnIn2S4 cluster; e, f TEM images of optimal sample (TiO2@ZnIn2S4 hollow nanospheres)

Full size image

The time-dependent water splitting over the as-prepared samples were tested under overall light irradiation. As shown in Fig. 4a, the pure TiO2 and ZnIn2S4 cluster displays lower photocatalytic H2 evolution of 205 and 1859 μmol·g−1 within 4-h photoreaction. Interestingly, the TiO2/ZnIn2S4 cluster shows relatively higher H2 evolution activity of 4518 μmol·g−1 compared to the pristine samples. Incredibly, the optimal TiO2@ZnIn2S4 hollow nanosphere possesses the highest H2 evolution of 4958 μmol·g−1, which is 1.1 times higher than the TiO2/ZnIn2S4 cluster during the 4-h HER reaction. The photocatalytic recyclability of the optimal sample was measured under the same condition. As shown in Fig. 4b, the optimal sample maintains a promising HER activity from 4958 to 3560 μmol·g−1 after 4-cycle photoreaction. The above decline in the photocatalytic H2 activity may be due to the fact that the catalyst was consumed during the washing and drying process. On the other hand, the core–shell hollow structure has low mechanical strength and is easy to break, leading to the above result. In addition, the apparent quantum yield (AQY) of the TiO2@ZnIn2S4 hollow nanosphere was measured at the incident light (435 nm) through the use of band-pass filters under the photocatalytic reaction [14, 15]. The apparent quantum yield of the TiO2@ZnIn2S4 hollow nanosphere is 1.84%.

Fig. 4
figure 4

a H2 evolution for as-synthesized samples under overall light irradiation; b cycling runs experiments of photoinduced hydrogen evolution for optimal sample under overall light irradiation

Full size image

The efficient transfer and separation of charge carriers over the samples (TiO2 hollow nanosphere, ZnIn2S4 cluster, TiO2/ZnIn2S4 cluster and TiO2@ZnIn2S4 hollow nanosphere) were elucidated by transient photocurrent–time measurements. As shown in Fig. 5a, it is clearly observed that the optimal TiO2@ZnIn2S4 hollow nanosphere exhibits an apparently stronger photocurrent density compared to the other samples. Additionally, both the ZnIn2S4 cluster and TiO2/ZnIn2S4 cluster display relatively higher photocurrent density in comparison with the pure TiO2. Figure 5b shows electrochemical impedance spectra (EIS) Nyquist plots of TiO2@ZnIn2S4 hollow nanosphere. The impedance is a complex number that can be expressed as the real part and the imaginary part, where the horizontal axis Z′ represents the real impedance (Zre) and the vertical axis Z″ is imaginary impedance (Zim). It can be seen that the obtained composite photocatalyst represents the smallest arc radius among these samples, indicating that the nano-sized hollow structure has the lower internal resistance of charge transfer in comparison with the other TiO2-based nanoclusters.

Fig. 5
figure 5

a Transient photocurrent response spectra and b EIS Nyquist plots for as-synthesized samples; c surface photovoltaic spectrum of and d UV–Vis DRS spectra of as-synthesized samples; N2 adsorption/desorption isotherms of e TiO2/ZnIn2S4 cluster and f TiO2@ZnIn2S4 hollow nanosphere

Full size image

In addition, the SPS was used to further explore the efficiency of photogenerated electrons migration over the samples. To be extended, the stronger the SPS signal is, the easier the charges separation is. Figure 5c reveals that the TiO2@ZnIn2S4 hollow nanospheres show much stronger SPS response signal than that of TiO2/ZnIn2S4 cluster in the region of 300–450 nm, which clearly confirm the enhanced capability of charges separation, thus ultimately causing the improvement in HER activity. And the data of SPS spectra are consistent with the result of transient photocurrent test.

As shown in Fig. 5d, the optimized TiO2@ZnIn2S4 hollow nanosphere has the increased light absorption in the wave region of 300–450 nm among the as-formed samples. This phenomenon can be ascribed to the multiple scattering effect produced by the inner cavity of the hollow nanostructure, which is beneficial for the enhancement of the light absorption. More significantly, the light harvesting of TiO2/ZnIn2S4 cluster in the range of 300–450 nm is higher than that of the pure TiO2 and ZnIn2S4. And the pure TiO2 shows the lowest visible-light absorption in the range of 360–600 nm among the as-prepared samples. As evident from these results, it can be concluded that the hollow nanostructure has a positive influence on the optical properties of photocatalyst [16, 17]. As illustrated in Fig. 5e, f, both the TiO2/ZnIn2S4 cluster and the optimal TiO2@ZnIn2S4 hollow nanosphere sample show the typical type IV isotherm with the H3-type hysteresis loop. The detailed texture properties of the two samples are shown in Table 1, it can be seen that TiO2/ZnIn2S4 cluster exhibits low active surface area (11.18 m2·g−1), and TiO2@ZnIn2S4 hollow nanosphere has much larger active surface area (79.72 m2·g−1), which demonstrates that the nano-sized hollow structure could expose numerous active sites and produce relative larger porous structure for enhancing mass transfer and boosting photocatalytic H2 evolution reaction.

Table 1 BET surface area (SBET), average pore size (d) and pore volume (V) of as-prepared samples
Full size table

According to above studies, we attempt to explore the detailed mechanism of charge carrier migration for solar-driven hydrogen generation in the TiO2@ZnIn2S4 hollow nanosphere photocatalyst under overall light irradiation. It can be seen in Fig. 6 that due to the accurate positions of valence band (VB) and conduction band (CB) between TiO2 hollow nanosphere and ZnIn2S4 cluster, the type-II binary heterojunction can be easily formed. Therefore, when the TiO2@ZnIn2S4 sample was irradiated by the solar energy, both of them would be excited to generate electron–hole pairs. And the photocatalytic electrons of the ZnIn2S4 could transfer to TiO2 via the core–shell interface due to the formation of impacted junction between TiO2 and ZnIn2S4, and then, the massive electrons would reduce the protons in the interior cavity to produce hydrogen. Furthermore, the photoinduced holes on the VB of TiO2 can fast transfer to the VB of ZnIn2S4, and would be immediately consumed by the sacrificial agent of lattice acid [14, 18,19,20]. It is emphasized that the core–shell hollow structure would trigger the confinement effect and multiple scattering effect, which could further greatly enhance the HER activity under overall light irradiation [21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36].

Fig. 6
figure 6

Schematic illustration of photocatalytic mechanism of H2 generation process therein

Full size image

4 Conclusion

To sum up, a novel core–shell TiO2@ZnIn2S4 hollow nanostructure was successfully designed and synthesized via templating method assisted by hydrothermal synthesis process, and the formed type-II heterojunction in the as-prepared sample could significantly improve the separation and migration of charge carriers. Additionally, the novel core–shell hollow structure could provide suitable interior cavity to trigger the confinement effect and multiple scattering effect, both ultimately resulting in a great photocatalytic H2 generation of the obtained sample under overall light irradiation.