Introduction

The photocatalytic performance is strongly dependent on the catalyst architectures including geometry, morphology and hierarchical structures (Xiao et al. 2004). Recently, two-dimensional (2D) nanomaterials with characteristics of intrinsic quantum confined electrons and more specific surface areas have gained more and more attention in catalysis (Guo et al. 2015). For example, ultrathin C3N4/Bi2O5I2 layered nanojunctions display great photocatalytic activities (Xia et al. 2016). However, these 2D nanomaterials share the common limitations of high tendency to aggregate and restack due to the lack of intrinsic driving force for 2D anisotropic growth (Zhang and Xie 2013). The development of three-dimensional (3D) architectures assembled from 2D nanomaterials is an efficient strategy to cope with this challenge (Wang et al. 2014, 2015). Three-dimensional (3D) structured materials assembled from 2D nanomaterials can combine the advantages of the individual properties of their building blocks as well as the novel characteristics and properties of the secondary architectures (Li et al. 2016; Wu et al. 2013b). For example, they possess porous structures made from their adjacent nano-units, large specific surface area to supply abundant active sites (Liu et al. 2015), cavity-mirror effect to improve optical irradiation and the easy solid/liquid separation (Wu et al. 2013a; Wei et al. 2014), all of which lead to enhanced light-harvesting capacity. For instance, 3D Ni3S4 frames have both high free volume and high compressive strength compared to flat Ni3S4 sheets (Wang et al. 2015). Graphene sheets crumple into paper ball-like structures to make them aggregation-resistant even after mechanical compression (Luo et al. 2013). To date, metal chalcogenide materials have attracted significant attention in the photocatalysis field due to their low cost and specific optical, acoustic, electronic properties (Liu et al. 2015; Meng et al. 2016). Therefore, the preparation of 3D architectures from the assembly of 2D metal chalcogenide nanomaterials may lead to the formation of high performance photocatalyst (Sun et al. 2016).

As an important class of transition metal chalcogenides, Indium sulfide (In2S3) is of great interest due to its simple synthetic procedure, low toxicity, and potential applications in photocatalysis (Nayak et al. 2014). Typically, β-In2S3 is n-type semiconductor with defected spinel structure and a band gap of 1.9-2.3 eV (Zhou et al. 2014), making it a great candidate for photocatalytic applications (Nayak et al. 2014). Many efforts have been devoted to synthesize β-In2S3 with various architectures: 3D hierarchical-like β-In2S3 hollow microspheres (Rengaraj et al. 2011), walnut-like In2S3 microspheres (Chai et al. 2012) and β-In2S3 nanotubes (Liu et al. 2011) using microwave irradiation (Patra et al. 2006), solvothermal (Chai et al. 2012) and sonochemical (Gorai and Chaudhuri 2005) methods. However, most of the aforementioned processes depend on poisonous organic solvents as the reaction media, various surfactants as templates and highly toxic sulfur sources to control the ions activity (Lee et al. 2005; Du et al. 2008; Abdelhady et al. 2013). For example, the most commonly used sulfur sources for the synthesis of β-In2S3 are sulfur powder (Park et al. 2006), sodium sulphide (Chai et al. 2012), and lauryl mercaptan (Liu et al. 2011). Therefore, the development of alternative simple and green method is highly desired. Recently, novel inorganic functional materials are prepared using inorganic-organic hybrids as single-source precursors. This method can combine the superior performance of both inorganic building blocks and organic components. Most importantly, inorganic-organic hybrids can serve as handy and effective precursors to prepare novel materials with good control of different components and the retention of macro-morphology (Wu et al. 2015). This single-source precursor route has several appealing features. First of all, it offers the potential advantages of mildness and simplified fabrication procedure and does not need exact control over starting material stoichiometry. Moreover, unusual crystal growth selectivity or metastable phase formation may be achieved, which are sometimes unattainable with the conventional synthetic technique (Zhang et al. 2005). For instance, inorganic hollow CdχZn1-χSe nanoframe (Wu et al. 2012), and porous MoS2 nanotubes (Zhuo et al. 2013b) have been synthesized through an ion-exchange and component stripping strategy using inorganic-organic hybrids.

Herein, we reported a facile hydrothermal strategy to synthesize hierarchical flower-like β-In2S3 assembled by 2D thin nanosheets. This is a green synthetic method which distilled water as the only solvent, and does not require post-treatment procedure. The flower-like hierarchical structures were prepared by decomposition of single-source precursor indium diethyldithiocarbamate (In-DDTC). The prepared hierarchical flower-like β-In2S3 possessed large surface area, good adsorption performance for cationic dye of methylene blue (MB) and great catalytic properties for the decolorization of methyl orange (MO) and tetracycline (TC). To the best of our knowledge, there have been very few reports on the application of In2S3 material as efficient photocatalyst for the removal of antibiotics in wastewater.

Experimental procedure

Materials

Indium nitrate hydrate (In(NO3)3 .xH2O) was purchased from Aladdin Industrial Corporation. Sodium diethyldithiocarbamate trihydrate (C5H10NNaS2 .3H2O) and ethyl alcohol were purchased from Sinopharm Chemical Reagent Co. Ltd., and all the analytical reagents were used without further purification.

Synthesis of hierarchical flower-like β-In2S3

Synthesis of indium diethyldithiocarbamate complex

In a typical procedure, 0.002 mol In(NO3)3 .xH2O and 0.006 mol sodium diethyldithiocarbamate were dissolved in 40 mL deionized water with agitation. The mixed solution was formed a stable white suspension and stirred for 2 h. The white product was collected by filtration, washed with deionized water and dried in air at 60°C.

Synthesis of hierarchical flower-like β-In2S3

Subsequently, a certain amount of the as-prepared single-source precursors were added into a 50 mL PTFE-lined stainless steel autoclave containing 40 mL deionized water, and the pH of the solution was adjusted to 3.0 by adding diluted nitric acid. The autoclave was sealed, maintained at 180°C for 12 h. After cooling to room temperature naturally, the obtained yellow precipitate was centrifuged at 8000 rpm for 5 min, washed with ethanol and deionized water for several times and dried overnight at 60°C. The control experiments were conducted by following similar procedure.

Results and discussion

Crystal structure and morphology

The hierarchical flower-like β-In2S3 assembled by 2D nanosheets was yellow powder (inset of Fig. 1a). The crystalline phase and morphology of the In2S3 were characterized by XRD, SEM and TEM. The typical XRD pattern of the nanostructures was shown in Fig. 1a. The peaks related to the (440), (211), (311), (400) planes could be indexed as cubic β-In2S3 phase by careful comparison with JCPDS card file no. 32-0456. No other impurities such as In2O3 or In(OH)3 were detected, indicating the high purity the sample, which were consistent with description in the literature (Chen et al. 2008). The powder had low crystallinity, which might show unique physical and chemical properties with more active sites and isotropic nature (Zhang et al. 2015).

Fig. 1
figure 1

XRD pattern, inset is the photograph of as-prepared the hierarchical flower-like β-In2S3.

Full size image

The morphology of the as-prepared In2S3 was investigated by SEM and TEM. As could be seen in Fig. 2 (a-b), the irregular flower-like spherical architectures of In2S3 were constructed from many interconnected thin nanosheets with thickness of ca. 5-7 nm. Notably, the flower-like β-In2S3 architectures possessed a number of mesoporous, which would help to increase the contact area and accelerate the ion transfer. The more detailed structural information of these hierarchical submicrospheres was revealed via TEM and HRTEM. Fig. 2 (c-e) proved that the submicrospheres were built from numerous small thin 2D nanosheets, supporting the SEM analysis (Fig. 2b). The HRTEM image of a single nanosheet (Fig. 2f) revealed that the lattice fringes were perfectly aligned across the surface. The measured interplanar spacing was 0.3101 nm, corresponding well with the (400) plane of cubic In2S3.

Fig. 2
figure 2

(a-b) SEM images; (c-e) TEM images; (f) HRTEM image of as-prepared the hierarchical flower-like β-In2S3.

Full size image

To better understand the fabrication process for the hierarchical flower-like β-In2S3 assembled by 2D thin nanosheets, and time-dependent evolution of morphology at 180°C was elucidated by SEM (Fig. 3a-f). After solvothermal treatment for 2 h, a large amount of prisms appeared, with sizes ranging from several hundred nanometers to a few micrometres (Fig. 3a). There were also small amounts of particles with curved surface (Fig. 3b). After 4 h, more flower-like hierarchies were formed few larger prisms were observed (Fig. 3c). The surface of large prisms grew coarse, indicating the beginning of the change of morphology. After 6 h, prisms were all already transformed into flowery hierarchies composed of mutual crisscross nanosheets (Fig. 3e). Meanwhile, a solid with folded surface could be found, which might be the intermediate in the process of morphological evolution (Fig. 3f). After 12 h, the submicrospheres were formed from numerous small thin 2D nanosheets, along with some imperfect morphology (Fig. 2a-e). Clearly, this was a gradual, three-stage morphological evolution process: pre-crystallized nuclei and crystal growth of primary particles; oriented attachment of prisms; and Ostwald-ripening of flowery hierarchies (Fig. 3g). At high temperature and vapor pressure, the precursor decomposed and quickly formed pre-crystallized nuclei and crystal growth into prisms. Due to the anisotropic structure of crystal, the prisms further grew into 2D nanosheets through oriented attachment. Furthermore, under longer hydrothermal treatment, large flowery hierarchies might collapse and generate some rupture parts of small flowers. These kinds of flowery hierarchies provided a large active surface area and higher efficiency of incident photons, and leading to higher photocatalytic activity. The final stage for the formation of flower-like structure could be due to the Ostwald ripening. The more accurate formation mechanism is currently under investigation.

Fig. 3
figure 3

SEM images of the samples collected at different time period: (a)-(b) 2h; (c)-(d) 4h; (e)-(f) 6h; (g) Schematic illustration of the formation mechanism of the flower-like structure.

Full size image

Composition and chemical states

EDX analysis was performed on the as-prepared hierarchical flower-like β-In2S3 to identify the elements present and measure their composition. The EDX spectrum (Fig. 4a) confirmed that there was no element other than In and S presented in the sample. The chemical state of the as-synthesized sample was further characterized by XPS analysis. The typical survey spectrum of the hierarchical flower-like β-In2S3 was shown in Fig. 4b. It revealed that no peaks of other elements except In, S, O, and C were observed. The peaks of C and O came from the reference sample and adsorbed oxygen (Zhou et al. 2014). The atomic ratio of [S] : [In] was estimated to be 1.43 from the survey spectrum confirming In2S3. The result was little smaller than the theoretical value of 1.5, which indicated the existence of sulfur vacancies or oxidation on the as-synthesized In2S3 sample surfaces (Tian et al. 2013). Fig. 4(c-d) showed In 3d and S 2p high-resolution region spectra, respectively. The peaks at the binding energy value of 444.8 and 452.4 eV were related to In 3d5/2 and In 3d3/2, while the peaks at 161.4 and 162.4 eV could be attributed to S 2p3/2 and S 2p1/2 transitions (Nayak et al. 2014; Gao et al. 2015b), respectively. The observed binding energy values of In 3d and S 2p agreed well with the reported data for In2S3 (Chai et al. 2012). To further study the O state on the surface of the In2S3, the high resolution O 1s XPS spectrum was analyzed and exhibited in Fig. 5 . As shown in Fig. 5 , the O 1s peak of In2S3 could be deconvoluted into three peaks. The peak located at around 530.8 eV, which was indexed to the oxygen from the crystal lattice (O2−). The other two peaks at 531.6 eV and 532.8 eV were referred to surface hydroxyl oxygen of adsorbed water and adsorbed oxygen from the ambient atmosphere, respectively (Du et al. 2009).

Fig. 4
figure 4

XPS region spectra of the hierarchical flower-like β-In2S3: (a) EDX spectrum; (b) XPS survey spectrum; (c) In 3d ; (d) S 2p.

Full size image
Fig. 5
figure 5

O 1s region XPS spectra of the In2S3 sample

Full size image

Porous structure determination by N2 adsorption / desorption

The specific surface area and porosity of the flower-like β-In2S3 hierarchical structures assembled by 2D thin nanosheets were further investigated by the nitrogen adsorption-desorption method. Fig. 6a showed the typical sorption isotherms and the corresponding pore size distribution (inset of Fig. 6a) of the flower-like β-In2S3. The type-IV isotherm with a hysteresis loop was obtained which could be identified as characteristic of mesoporous materials. These mesopores arose from the spaces among the nanosheets, consistent with the result of the HRTEM measurement. Accordingly, the pore size distribution curve depicted a wide range of pore size distribution from 2 to 150 nm with a unimodal shape at 32 nm. The BET surface area of as-prepared the hierarchical flower-like β-In2S3 was calculated to be 134.1 m2 g-1. The as-prepared hierarchical flower-like β-In2S3 showed the higher surface area among various previously reported In2S3 nanomaterials (Table 1). The large surface area would facilitate the surface adsorption of reactants and promote interfacial charge transfer (Zhou et al. 2013), and improve the photocatalytic properties.

Fig. 6
figure 6

As-prepared hierarchical flower-like β-In2S3: (a) N2 adsorption/desorption isotherm curves (inset: BJH pore size distribution plot); (b) UV-vis DRS spectrum.

Full size image
Table 1 the summary of reported BET surface area of various In2S3 materials
Full size table

DRS property

The optical absorption properties of a semiconductor were relevant to the electronic features, and recognized as a key factor in determining it photocatalytic activity. Fig. 6b showed the UV-vis diffuse reflectance spectra (DRS) of the hierarchical flower-like β-In2S3. The band gap (Eg) of In2S3 was reported to vary between 2.0 and 2.2 eV, which corresponded to 620-550 nm (Rengaraj et al. 2011). The absorption edge about 580 nm was evidently seen, demonstrating that the band gap of the product was about 2.14 eV. Moreover, the hierarchical flower-like β-In2S3 had strong absorption ranging from the visible to UV region, suggesting great effect of light absorption.

Adsorption performance

To investigate the surface charge of as-prepared sample, zeta potential measurement was carried out. The hierarchical flower-like β-In2S3 had a low overall charge with a zeta potential of about -24.2 mV, which could be due to the adsorption of -OH on the In2S3 surface in aqueous solution. This could promote efficient adsorption through providing adsorption sites for interaction with the cationic groups of MB.

Adsorption kinetics studies were explored at different time interval of MB (35 mg L-1) and adsorbent (10 mg) interaction. Fig. 7a showed adsorption kinetic of MB on In2S3 at 298 K. Both pseudo-first-order and pseudo-second-order kinetics were tested using the experimental data of MB removal from aqueous solutions. The pseudo-first-order rate equation is given by the following equation (Mittal et al. 2009):

$$ \ln \left({\mathrm{Q}}_{\mathrm{e}}-{\mathrm{Q}}_{\mathrm{t}}\right)= \ln \left({\mathrm{Q}}_{\mathrm{e}1}\right)-{k}_1\mathrm{t} $$
(1)
Fig. 7
figure 7

(a) Adsorption kinetic (T = 298 K; C0 = 35 mg L-1; amount of adsorbent = 10 mg; volume of solution = 50 mL); (b) Pseudo-first-order kinetics; (c) Pseudo-second-order kinetics; (d) Adsorption isotherm (T=298 K; adsorbent dose =10 mg; dye concentration = 30-60 mg L-1; solution volume = 50 mL); (e) Langmuir isotherm; (f) Freundlich isotherm.

Full size image

where Qe and Qt are the amounts of dye adsorbed at equilibrium (mg g−1) and at contact time t (min), respectively. Qe1 and k 1 show the theoretical equilibrium adsorption and rate constant of pseudo-first-order kinetic, respectively. The values k 1 and Qe1 could be determined from the slope and the intercept of the plot of ln(Qe − Qt) against t and presented in Fig. 7b. The rate of pseudo-second-order model depend on the amount and the quantity of dye adsorbed on the surface of adsorbent (Gucek et al. 2005). The pseudo-second-order rate can be expressed as follows (Gupta et al. 2010):

$$ \mathrm{t}/{\mathrm{Q}}_{\mathrm{t}}=1/{k}_2{\left({\mathrm{Q}}_{\mathrm{e}2}\right)}^2+\mathrm{t}/{\mathrm{Q}}_{\mathrm{e}2} $$
(2)

where k 2 (g mg−1 min−1) is the pseudo-second-order rate constant, and Qe2 is the theoretical adsorbed dye (mg g−1). In this case, the slope and the intercept of the plot of t/Qe vs t gave k 2 and Qe2 values as presented in Fig. 7c. As it could be seen from Table 2, the high R2 value suggested that the pseudo-second-order kinetic model was more suitable than pseudo-first-order kinetic model for the MB removal. Moreover, the Qe2 value for the second-order kinetic was close to the experiment (Qe,exp). This behavior could be explained with the existence of chemisorptions (Saha et al. 2012), which was in accordance with literature result (Madaeni et al. 2011).

Table 2 Summary of the kinetic constants obtained from linear regression of the two models
Full size table

To further investigate the adsorption mechanism, Langmuir and Freundlich models were used to describe the equilibrium of adsorption. Fig. 7d showed adsorption isotherm of MB on In2S3 at 298 K. It is known that the Langmuir model predicts the formation of a monolayer of the adsorbate on the adsorbent surface, while the Freundlich isotherm is an empirical equation for the description of equilibrium on heterogeneous surfaces. The isotherm linear forms of Langmuir and Freundlich are expressed by Eqs. (3) and (4), respectively:

$$ {\mathrm{C}}_{\mathrm{e}}/{\mathrm{Q}}_{\mathrm{e}}={\mathrm{C}}_{\mathrm{e}}/{\mathrm{Q}}_{\mathrm{m}}+1/\left({\mathrm{K}}_{\mathrm{L}}{\mathrm{Q}}_{\mathrm{m}}\right) $$
(3)
$$ \log {\mathrm{Q}}_{\mathrm{e}}= \log {\mathrm{K}}_{\mathrm{F}}+\frac{1}{\mathrm{n}} \log {\mathrm{C}}_{\mathrm{e}} $$
(4)

Where Qe is the amounts of dye adsorbed at equilibrium (mg g−1), Ce is the final concentration in solution of dye adsorbed at equilibrium (mg L-1), KL is the Langmuir constant (L mg−1), depending upon the adsorption energy, and Qm is the maximum adsorption capacity (mg g−1). KF (mg g−1) is the adsorption capacity of the adsorbent and n (L mg−1) is the Freundlich constant. Ce/Qe against Ce and logQe against logCe were plotted shown in Fig. 7e and f. And the data were analyzed by linear regression, the values of Qm and KL, n and KF calculated from the slope and intercept, respectively.

As it can be seen from Table 3, the Langmuir model showed the best fit (R2 = 0.9999) with a saturated coverage of 195.3 mg g-1, which was close to the experimental data (195.5 mg g-1). Therefore, monolayer adsorption was concluded and the adsorption sites were homogeneous.

Table 3 Correlation coefficients and isotherm rate constants for Langmuir and Freundlich models
Full size table

Photodecolorization performance

To evaluate the photoactivity of catalysts, methyl orange (MO), a very stable azo dye, was used as a model pollutant (An et al. 2013). UV-vis absorption measurement was a very simple and direct method to study small molecule substances with subtle changes in molecule structure. Without In2S3, almost no color removed of the MO after 4 h of irradiation (Fig. 8a). Before the irradiation, minimum decrease in MO concentration was observed after stirring the mixture of MO and In2S3 for 30 min in the dark condition (Fig. 8b). As could be observed from these spectra, the maximum absorption located at 465 nm and 270 nm. The peak at 465 nm was originated from an extended chromophore___ azo linkage and the peak at 270 nm was associated with benzene ring in the molecule, respectively (Zhang et al. 2006). After irradiation, the intensity of the two peaks disappeared after 15 min of irradiation, indicating the destruction of its chromophoric structure (Fig. 8c). The photocatalytic decolorization kinetics of MO were shown in Fig. 8d, although the ratio of C/C0 decreased with the illumination time was quickly, the ratio of TOC/TOC0 decreased with the illumination time was slowly. These results indicated that MO was successfully decolorized to a colorless state by In2S3, but the MO molecules could not be fully photomineralized. As reported in several literatures, the intermediates were stable in the system which required a longer time for further oxidation (Nur Farhana et al. 2012; Yu et al. 2016). The MS analysis was performed for MO decolorization, and the formation of intermediates during the decolorization process was identified by this technique. The sample before the MO decolorization showed one big peak at m/z 304. In contrast, the sample after 15 min of decolorization showed one main m/z peak at 172, corresponding to 4-aminosulfonic acid (Han et al. 2016). The results suggested that decolorization proceeds through the cleavage of azo group —N=N— connecting the two aromatic rings to amines, which agreed with literature results (Parshetti et al. 2010). Clearly, the decoloration of MO was attributed to the photocatalytic decolorization instead of adsorption.

Fig. 8
figure 8

Time-dependent UV-vis absorption spectra of MO: (a) Without β-In2S3, (b) With β-In2S3 under dark condition, and (c) With β-In2S3 under visible light; (d) Photocatalytic degradation kinetics of MO.

Full size image

We further studied the durability and the recyclability of the hierarchical flower-like β-In2S3 for the decolorization of MO. The decolorization efficiency was maintained at 98.8% after the sixteenth cycle (Fig. 9c), indicating a negligible change in decolorization performance and high stability of the catalyst. After the sixteenth cycle of the decolorization experiment, the catalyst powder was collected by centrifugation and dried at 60 °C for 4 h. The phase of used catalyst was remained unchanged (Fig. 9d).

Fig. 9
figure 9

(a) In2S3 catalyst activity of capture agent; (b) Proposed mechanisms of photocatalytic reaction in In2S3 system, the nets structure is In2S3 nanosheet; (c) Photocatalytic decolorization curves of MO under visible light irradiation the first and the sixteenth cycles; (d) XRD patterns of as-prepared In2S3 before and after 16 cycles.

Full size image

Different pathways were reported for the visble-light-driven photodecolorization of MO, which could be classified according to the active species (An et al. 2013; Gao et al. 2015a). Trapping experiments were conducted to determine the main active species during the photodecolorization process. Benzoquinone (BQ), isopropyl alcohol (IPA), EDTA-2Na, and K2Cr2O7 were used as scavengers of superoxide radical (.O2 -), hydroxyl radical (.OH), h+ and e-1 (Song et al. 2016). As shown in Fig. 9 a, the decolorization of MO molecules was attributed to the predominant action of oxidation action of the generated .O2 − radicals and subordinate e-1 reduction process. Hydroxyl radical was confirmed to be insignificant for the decolorization of MO (Li et al. 2011). The CB and VB potentials (ECB and EVB) of the β-In2S3 can be calculated by the empirical equations of ECB = χ − Ee − 0.5Eg and EVB = ECB + Eg, where Eg = 2.14 eV, χ is the electronegativity of the In2S3 (4.69 eV), Ee is the energy of free electrons on the hydrogen scale (about 4.5 eV) (Chen et al. 2016). Therefore,the value of ECB and EVB are -0.88 and 1.06 eV for the β-In2S3. The valance band of In2S3 was more negative than E (.OH/OH−) (2.38 V vs. NHE) and E (.OH/H2O) (2.27 V vs. NHE), hence the photoinduced holes could not oxidize OH− and H2O to produce .OH. Instead, they could oxidize the dye macromolecules into directly micromolecules and final products (Cheng et al. 2010). While the conduction band of In2S3 was more negative than E (O2/.O2 −) and the photogenerated electrons of the In2S3 sample were enough to generate .O2 − under visible irradiation. The schematic demonstration of the excitation and charge transfer processes of the hierarchical flower-like β-In2S3 under visible light irradiation was shown in Fig. 9b. The proposed mechanism for the photocatalytic decomposition of MO by the flower-like β-In2S3 assembled by 2D thin nanosheets can be described as follows:

$$ \begin{array}{cc}\hfill {\mathrm{In}}_2{\mathrm{S}}_3+\mathrm{h}\upnu \to {\mathrm{In}}_2{\mathrm{S}}_3\left({\mathrm{e}}^{\hbox{-} 1}+{\mathrm{h}}^{+}\right)\hfill & \hfill \hfill \\ {}\hfill {\mathrm{O}}_2+2{\mathrm{H}}^{+}+2{\mathrm{e}}^{\hbox{-} 1}\to {\mathrm{H}}_2{\mathrm{O}}_2\hfill & \hfill {\mathrm{E}}^{\uptheta}=0.6824\ \mathrm{V}\hfill \\ {}\hfill {\mathrm{O}}_2+{\mathrm{e}}^{\hbox{-} 1}\to {{\mathrm{O}}_2}^{.\hbox{-}}\hfill & \hfill {\mathrm{E}}^{\uptheta}=\hbox{-} 0.563\ \mathrm{V}\hfill \\ {}\hfill {{\mathrm{O}}_2}^{.\hbox{-} }+{\mathrm{H}}_2\mathrm{O}+{\mathrm{e}}^{\hbox{-} 1}\to {{\mathrm{H}\mathrm{O}}_2}^{\hbox{-} }+{\mathrm{O}\mathrm{H}}^{\hbox{-}}\hfill & \hfill {\mathrm{E}}^{\uptheta}=0.413\ \mathrm{V}\hfill \\ {}\hfill {{\mathrm{H}\mathrm{O}}_2}^{\hbox{-} }+{\mathrm{H}}_2\mathrm{O}+{\mathrm{e}}^{\hbox{-} 1}{\to}^{.}\mathrm{O}\mathrm{H}+2{\mathrm{O}\mathrm{H}}^{\hbox{-}}\hfill & \hfill {\mathrm{E}}^{\uptheta}=0.878\ \mathrm{V}\hfill \\ {}\hfill {\mathrm{h}}^{+}/{\mathrm{H}}_2{\mathrm{O}}_2{/}^{.}\mathrm{O}\mathrm{H}+\mathrm{MO}\to \mathrm{degraded}\ \mathrm{or}\ \mathrm{mineralized}\ \mathrm{products}\hfill & \hfill \hfill \end{array} $$

The hierarchical flower-like β-In2S3 assembled by 2D nanosheets was also a highly effective photocatalyst for antibiotics removal under visible light irradiation. The catalytic degradation of antibiotics was of great importance because they posed serious threats to the ecosystem and human health even at small concentrations in water system. As shown in Fig. 10, TC could hardly be decolored without any photocatalyst (Fig. 10a), while the flower-like β-In2S3 showed high activity with decolorization rate of 90.1 % after visible light irradiation for 60 min. The decolorization speed of TC was much faster than previously reported systems (Wang et al. 2016; Ai et al. 2015).

Fig. 10
figure 10

(a) Time-dependent UV-vis absorption spectra; (b) Photocatalytic decolorization curves of TC under visible light in the presence of the hierarchical flower-like β-In2S3.

Full size image

The as-synthesized hierarchical flower-like β-In2S3 occurred a rapid decolorization reaction that may be due to the following several reasons for synergism. The first reason was attributed to its cubic phase crystal structure. Naik et al. (2008) reported that the cubic In2S3 showed higher photoactivity for hydrogen production than of tetragonal β-In2S3. Dai et al. (2010) proposed that cubic In2S3 had not the three-dimensional defect structure resulted from the ordering of indium vacancies, hence the charges transportation efficiency of the pathway would not be depressed. Meanwhile, sulfur vacancies in the as-synthesized In2S3 surfaces might exhibit electron affinity and could act as electron traps which promoting the separation of photogenerated charge carriers (Gao et al. 2015b). Moreover, the sample contained oxygen from the crystal lattice was more propitious to form the defect band (Chen et al. 2016). Second, not only the high surface area but also the wide range of pore size distribution of the flower-like In2S3 led to greater contact possibility with pollutants and provided more active catalytic sites (Dai et al. 2010; Li et al. 2014). Third, the smaller size of the flower-like β-In2S3 hierarchical structures than other flower-like β-In2S3 microspheres (Wei et al. 2014; Liu et al. 2011) (> 2 μm) could effectively shorten the diffusion distance for carriers (Maeda 2011; Mi et al. 2015). Finally, the photoexcited carriers separated and migrated to the surface with less recombination as a result of the ultrathin sheet (Mi et al. 2015).

Conclusion

In summary, a simple and facile method was developed to transform single-source precursor In-DDTC into flower-like In2S3 hierarchical structures assembled by 2D thin nanosheets. The flower-like In2S3 material showed high adsorption property for MB and great decolorization property for MO and TC under the visible light irradiation. The decreased speed of the TOC was slower than the MO concentration. Several factors including its unique structures were proposed to be responsible for the great properties. This facile and promising synthetic strategy could be extended to prepare a wide variety of functional materials for potential applications including photocatalytic reactions, hydrogen production, solar cells and lithium-ion batteries.