Study on the relationship between Bi₂S₃ with different morphologies and its photocatalytic hydrogen production performance

Abstract

The morphology of a material is considered one of the primary aspects affecting its photocatalytic performance. Various methods have been developed to tailor the morphology of photocatalytic materials for photocatalytic water splitting. Bi₂S₃ is an excellent photoabsorption material with relatively narrow band gaps. Herein, Bi₂S₃ samples with different morphologies are successfully prepared via a simple one-step hydrothermal method and employed effectively as visible light-driven photocatalysts for hydrogen production. Electron microscopy technologies were used to characterize the morphology and microstructure of the Bi₂S₃ samples, which exhibit three kinds of morphologies, namely nanotubes, nanoflowers and nanorods. As a result, the Bi₂S₃ nanotubes have the largest BET specific surface area and lowest PL intensity, and these characteristics lead to having the best hydrogen production rate. Moreover, the catalysis mechanism is simply explained by studying the relationship between the morphology and microstructure of a material and its photocatalytic performance.

Introduction

Because of the energy shortage and the pressure of environmental protection, it is particularly important to develop new forms of energy other than fossil fuels. Photocatalytic hydrogen production has great potential because it is derived from natural sources such as water and solar energy, which are highly available, renewable and environmentally friendly (Suk et al. 2012; Qu et al. 2020). Since water splitting has been achieved on TiO₂ electrodes through a photoelectrochemical (PEC) approach to produce hydrogen, semiconductor technology for the photocatalytic decomposition of water has attracted the attention of many researchers (Fujishima and Honda 1972). Semiconductor catalysts play a vital role in the process of photocatalytic hydrogen production, both in the PEC reaction that was first discovered and the photochemical processes that followed (Liao et al. 2012). There are many factors affecting the performance of semiconductor photocatalysts, such as the band gap, structure and morphology, corrosion resistance, solution pH, and operating temperature (Ahmad et al. 2015; Chen et al. 2010; Maeda and Domen 2010). However, for the selection of photocatalytic materials, the band gap is the most critical factor because it determines the absorption and utilization efficiency of light energy. Unfortunately, traditional photocatalytic materials, such as TiO₂, ZnS and ZnO, are semiconductors that use only ultraviolet solar energy. Therefore, the development of photocatalytic materials with a wide spectral absorption range is of far-reaching significance to provide a breakthrough for photocatalytic hydrogen production applications (Liu et al. 2011; Chen et al. 2016; Ma et al. 2020).

Bi₂S₃ is a semiconductor with a relatively narrow band gap of 1.3–1.7 eV that can absorb visible light and even near-infrared light. It has been proven to be an excellent photoabsorption material widely used in photonic devices, such as photodetectors, solar cells, and all-optical diodes (Xu et al. 2012; Yu et al. 2019; Shan et al. 2019; Martinez et al. 2011). Its superwide light absorption range largely compensates for the shortcomings of TiO₂, WO₃, ZnS and other wide bandgap semiconductors as photocatalytic materials (Wang et al. 2017a; Kumar et al. 2016; Liu et al. 2021; Liu et al. 2016; Xiong et al. 2016). Sandeep Kumar et al. reported that TiO₂ nanoparticles were coupled with the Bi₂S₃ semiconductor to synthesize a heterostructure, which had a high visible light response. The results showed that the photocatalytic degradation efficiency of amaranth dye under solar light was much better than that of bare Bi₂S₃ and TiO₂ (Kumar et al. 2016). Zhu Liu et al. prepared a novel Bi₂S₃/BiVO₄/TiO₂ ternary heterostructure photocatalyst with a large light absorption coefficient. Compared with TiO₂ nanocrystals, the Bi₂S₃/BiVO₄/TiO₂ photocatalyst had a higher absorption intensity and extended the absorption range to near infrared. Bi₂S₃/BiVO₄/TiO₂ had a photocurrent of more than 10 times that of bare TiO₂ when performing photochemical experiments and had an approximately 4 times higher photocatalytic degradation efficiency than bare TiO₂ (Liu et al. 2021). Bi₂S₃/WO₃ thin films prepared by coating a layer of Bi₂S₃ on the surface of WO₃ nanoplate arrays exhibited higher PEC performance. The Bi₂S₃-modified WO₃ photoelectrode displayed a significantly higher photocurrent and higher electron transport rate due to the significant enhancement in response to visible light and good interfacial contact between the two crystals (Liu et al. 2016). A Bi₂S₃/ZnS composite was synthesized to obtain enhanced light absorption and a redshifted absorption edge for more efficient separation of light-generated electron–hole pairs. The photocatalytic degradation performance of hexagonal Bi₂S₃/ZnS composites for methylene blue was better than that of pure Bi₂S₃ and pure ZnS, and the degradation rates were 7.1 times and 3.6 times higher than those of pure Bi₂S₃ and pure ZnS, respectively (Xiong et al. 2016).

Regarding materials with certain chemical compositions, a large number of studies have shown that the photocatalytic performance is closely related to the crystalline structure and morphology characteristics. Jiarui Li et al. reported that hexagonal BiPO₄, which has abundant phosphate defects, exhibited stronger photocatalytic activity than pure BiPO₄. DFT calculations revealed that defects induced the formation of intermediate energy levels within the band gap, allowing the effective charge to be transferred from the valence band to the conduction band (Li et al. 2019). The effects of morphology adjustment and defect engineering on the corresponding photocatalytic activity of CaCu₃Ti₄O₁₂-degrading antibiotics were systematically investigated by Reshalaiti Hailili, who revealed that the optimal photocatalytic performance was attributed to its unique morphology and effective carrier separation due to local defects (Hailili et al. 2019). Various methods have been developed to tailor the morphology of photocatalytic materials or to deposit photocatalytic materials on substrate materials with a unique morphology because the influence of morphology on the photocatalytic efficiency of materials is widely known (Hailili et al. 2019; Phan and Shin 2011; Wang et al. 2018; Farhadian et al. 2015). Different morphologies of TiO₂ materials have been used for the photocatalytic degradation of methylene blue. Among previously reported photocatalytic results, TiO₂ particles with prismatic and flower-like morphologies showed the highest photocatalytic activity, which was due to the synergistic effect between the morphology and crystal miscibility of TiO₂ (Phan and Shin 2011). The metal–organic framework compound ZIF-67 was synthesized with different morphologies for CO₂ reduction. The photocatalytic performance of ZIF-67 showed that the leaf shape of two-dimensional ZIF-67 has the best photocatalytic activity and stability, which was due to its highest adsorption capacity for CO₂ and effective electron transfer (Wang et al. 2018). WO₃ nanostructures with nanorod, nanosphere and nanoplate morphologies were prepared and characterized in this study to investigate the influence of shape on photocatalytic performance. The results show that WO₃ nanoplates have the best photocatalytic activity because of their lower coordination number, and the atoms located at the edges and corners of WO₃ nanoplates are more active (Farhadian et al. 2015).

However, there is very little literature on the influence of the structure and morphology of pure Bi₂S₃ crystals on photocatalytic performance. In this paper, Bi₂S₃ nanomaterials with nanotube, nanoflower and nanorod morphologies were synthesized by a hydrothermal method, and the influence of their crystallinity and morphology on their photocatalytic hydrogen production performance was studied.

Materials and methods

Preparation of Bi₂S₃ samples with different morphologies (Yao et al. 2009)

First, 1.226 g of Bi(NO₃)₃∙5H₂O was dissolved in 5 mL of ethanol and stirred to form solution A. Second, 1.53 g of Na₂S∙9H₂O was dissolved in 10 mL of deionized water to form solution B. Third, 0.91 g of CO(NH₂)₂ was dissolved in 20 mL of deionized water to form solution C. Then, solution B was slowly added into solution A with magnetic stirring to form a black solution. Next, solution C was added to the black solution. The above solution was transferred to a 50 mL autoclave, and the autoclave was placed in an oven at 120 °C for 12 h at a constant temperature. The above solution was naturally cooled to room temperature, and a black precipitate was obtained by centrifugation. The solution was washed several times with ethanol to obtain Bi₂S₃ nanotubes.

Bi₂S₃ nanoflowers and nanorods were also synthesized through a similar synthesis method as that of Bi₂S₃ nanotubes, but the sulfur source of Na₂S∙9H₂O was replaced by Cs(NH₂)₂ and Na₂S₂SO₃, respectively.

Material characterization

The crystal phase of the samples was determined via X-ray diffraction (XRD, PANalytical X’ Pert Powder) using a Cu-Kα radiation source (λ = 1.54186 Å) in the 2θ range of 10–70° at a voltage and current of 40 kV and 40 mA, respectively. Raman analysis was characterized on a Raman spectrometer (LabRAM HR Evolution) using a solid-state laser with a wavelength of 532 nm as the excitation source. The morphology and structure of the samples were obtained by field-emission scanning electron microscopy (SEM, FEI Quattro S) and transmission electron microscopy (TEM, FEI Talos F200S). N₂ adsorption/desorption isotherms were analyzed with a specific surface and aperture analyzer (BELSORP-max-II) at 77 K to obtain the specific surface area and pore volume of the samples. The particle size of the samples was measured three times by dynamic light scattering (NanoBrook Omni, America). UV–vis diffuse reflectance spectroscopy (DRS) was performed with a UV–visible near-infrared spectrophotometer (Shimadzu, UV-3600) from 600 to 1200 nm with BaSO₄ as the reflectance standard to obtain the optical properties of the samples. Photoluminescence (PL) measurements were performed with a luminescence spectrophotometer (FLS1000) at an excitation wavelength of 335 nm. Time-resolved photoluminescence spectra (TRPL) were performed on a fluorescence spectrophotometer (FLS1000).

Photocatalytic performance evolution

The photocatalytic hydrogen evolution reaction was conducted in a glass container connected to a closed circulation and evacuation system (Beijing Perfect Light Company). The photocatalytic hydrogen production performance of the samples was evaluated under simulated sunlight irradiation with a 300 W Xe lamp. First, 0.5 M Na₂SO₃ and 0.5 M Na₂S·9H₂O were added as sacrificial agents to a glass container with 100 mL of deionized water; then, 20 mg of the catalyst was dispersed in the above solution. Next, the glass container with the above suspensions was degassed for 30 min under continuous and vigorous stirring to eliminate the gas and provide a vacuum environment for the reaction. The evolved H₂ of the samples was measured at 25 °C via gas chromatography (GC7900 Techcomp) with a thermal conductivity detector (TCD); Ar was used as the carrier gas.

Results and discussion

Structure and composition of Bi₂S₃ samples

To explore the crystal quality of Bi₂S₃ with different sulfur sources, the XRD patterns of the samples were analyzed, as shown in Fig. 1a. All the patterns of the samples exhibit sharp diffraction peaks, which indicate that all samples are highly crystalline. It can be seen that Bi₂S₃ exhibits orthorhombic crystal structure (JCPDS 17-0320) pattern, and no impurity peaks are detected, suggesting that all samples consist of a pure phase. Furthermore, the molecular structures of Bi₂S₃ were characterized by Raman spectroscopy, as shown in Fig. 1b. The peaks at 182, 237 and 258 cm⁻¹ correspond to the Ag, Ag₁ and B1g modes of Bi₂S₃ with different morphologies, respectively (Wang et al. 2017b). These results clearly confirm that Bi₂S₃ samples with different morphologies are successfully achieved.

Fig. 1

a XRD patterns and b Raman spectra of Bi₂S₃ with different morphologies

Morphological analysis of Bi₂S₃ samples

Bi₂S₃ prepared with different sulfur sources was characterized by SEM to reveal the variations in morphology and microstructure (Fig. 2a–f). The Bi₂S₃ nanotubes, based on the Na₂S∙9H₂O sulfur source, are composed of a large number of linear products with a smooth surface and uniform diameters and lengths of 1–5 μm; additionally, these nanotubes are hollow with an outer diameter of 60–140 nm, an inner diameter of 10–48 nm and a tube wall of approximately 40 nm (Fig. 2a and b). The Bi₂S₃ nanoflowers, based on the C_S(NH₂)₂ sulfur source, are similar to three-dimensional spherical polymers with diameters of 2–7 μm; additionally, these nanoflowers are made up of many flat bands of microcrystals, growing in a divergent form from the center to all sides (Fig. 2c and d). The Bi₂S₃ nanorods, based on the Na₂S₂SO₃ sulfur source, have diameters of approximately 200 nm and lengths between 2 and 4 nm (Fig. 2e and f). TEM was then performed on the Bi₂S₃ nanotubes, and the results are shown in Fig. 3. The material has a hollow structure and a smooth surface with an internal aperture of approximately 50 nm and an outer diameter of less than 200 nm, further confirming that the material is a nanotube structure (Fig. 3a and b). Furthermore, the lattice spacing of the material is determined to be 0.398 nm and 0.197 nm, which can be assigned to the (220) plane and (002) plane of Bi₂S₃, respectively (Fig. 3c and d).

Fig. 2

SEM images of the a, b Bi₂S₃ nanotubes; c, d Bi₂S₃ nanoflowers; and e, f Bi₂S₃ nanorods

Fig. 3

a, b TEM images of the Bi₂S₃ nanotubes and c, d HRTEM images of the Bi₂S₃ nanotubes

Surface area, pore structure and particle size of Bi₂S₃ samples

To identify the specific surface area and pore structure of the as-synthesized materials, N₂ adsorption–desorption isotherms were obtained for the Bi₂S₃ samples with different morphologies, as shown in Fig. 4. According to the IUPAC classification (Li et al. 2016), Bi₂S₃ nanotubes and Bi₂S₃ nanoflowers with a type IV isotherm represent a typical of mesoporous structures, while Bi₂S₃ nanorods with a type III isotherm show a nonporous structure. The specific surface area, pore size and pore volume analysis of the Bi₂S₃ samples are also presented in Table 1. The BET specific surface area of the Bi₂S₃ nanotubes is 14.5490 m²/g, which is larger than those of the Bi₂S₃ nanoflowers (2.4578 m²/g) and nanorods (7.9459 m²/g). Simultaneously, the pore size distribution curve of three Bi₂S₃ samples is displayed in Fig. 4b. The pore size and pore volume are not available for Bi₂S₃ nanorods with a nonporous structure. Moreover, there is abnormal distribution curve for the pore size distribution of Bi₂S₃ nanotubes and Bi₂S₃ nanoflowers. Therefore, the average pore diameter is not relevant to the maximum value and middle pore size of the dV/dpW. Average pore size is usually used to describe the pore distribution of the sample. The average pore diameter and pore volume of Bi₂S₃ nanotubes (47.768 nm and 0.1596 cc/g) is larger than that of Bi₂S₃ nanoflowers (35.995 nm and 0.0345 cc/g). Due to the largest BET specific surface area, larger pore size and larger pore volume, Bi₂S₃ nanotubes do not only have more channels to facilitate the transmission of substances, but also provide more space for intermediate products. A larger specific surface area, larger average pore size and larger mesoporous volume provide many favorable advantages, including favoring the adsorption of water molecules, facilitating the transport of photogenerated carriers and providing more reactive active sites, thus promoting the photocatalytic performance of the material (He et al. 2021; Du et al. 2021; Wang et al. 2020). As mentioned above, the low PL intensity may correlate with the efficient charge transfer in the Bi₂S₃ nanotubes with the largest BET surface area, pore size and pore volume, thereby confirming the prevention of the direct recombination of electrons and holes. The particle size of the samples was measured by dynamic light scattering. The average particle size of Bi₂S₃ nanotubes (467 nm) is smaller than that of Bi₂S₃ nanoflowers (596 nm) and Bi₂S₃ nanorods (896 nm). It can be inferred that smaller size of Bi₂S₃ nanotubes is favorable for photocatalytic H₂ evolution.

Fig. 4

a N₂ adsorption–desorption isotherms and b pore size distribution of the Bi₂S₃ samples with different morphologies

Table 1 Surface area, pore size and pore volume analysis of the Bi₂S₃ samples with different morphologies

Optical properties of Bi₂S₃ samples

The optical properties of the Bi₂S₃ catalysts with different morphologies are presented in Fig. 5. It is clearly shown that the as-prepared Bi₂S₃ has a good absorption region from the UV to infrared region; moreover, the Bi₂S₃ nanotubes demonstrate a distinct absorption band edge and no absorption after 1000 nm. In contrast, the edges of the absorption bands of the Bi₂S₃ nanoflowers and nanorods are not obvious (Fig. 5a). Furthermore, the PL intensity is positively correlated with the carrier combination rate. It is obvious that the PL emission intensity of the Bi₂S₃ nanotubes is lower than that of the Bi₂S₃ nanoflowers and nanorods (Fig. 5b). From the PL results, it can be deduced that the charge separation rate of Bi₂S₃ nanotubes is promoted under visible light irradiation. Simultaneously, TRPL spectra (Fig. 5c) were measured to demonstrate the charge carrier kinetics of Bi₂S₃ samples. As shown in Table 2, Bi₂S₃ nanotubes exhibit a longer average lifetime (3.0662 ns) than Bi₂S₃ nanoflowers (2.3734 ns) and Bi₂S₃ nanorods (2.1913 ns), indicating that charge separation efficiency is improved and the recombination of photoinduced electron–hole pairs is inhibited in the Bi₂S₃ nanotubes. These results reveal that more photoinduced electrons and holes at the interface of Bi₂S₃ nanotubes can contribute to a higher photocatalytic activity, as the charge recombination can be effectively inhibited.

Fig. 5

a UV–vis DRS spectra, b photoluminescence spectra and c time-resolved fluorescence emission spectra of the Bi₂S₃ samples with different morphologies

Table 2 The PL lifetime parameters of the Bi₂S₃ samples with different morphologies

Photocatalytic hydrogen production of Bi₂S₃ samples

The photocatalytic activity of the Bi₂S₃ samples was evaluated through the photocatalytic splitting of water under simulated solar light irradiation. The time-dependent H₂ yields of the Bi₂S₃ samples with different morphologies are displayed in Fig. 6a and b. Clearly, compared with the Bi₂S₃ nanoflowers and nanorods, the Bi₂S₃ nanotubes with their hollow tubular structure exhibit the best rate of hydrogen production (271 µmol/h/g). In addition, it can be seen from Fig. 6c and d that the Bi₂S₃ samples with different morphologies can be maintained for four cycles with a total reaction time of 8 h, suggesting their good stability during the photocatalytic hydrogen production process. The photoreduction activity of a material depends on many factors. The first is a large BET specific surface area, which usually implies a large number of active sites on the surface. Bi₂S₃ nanotubes have both internal and external passages, thereby exhibiting a larger specific surface area that has more active sites for the photocatalytic hydrogen production reaction. A low PL intensity is another important factor that indicates the low recombination of charge carriers. The Bi₂S₃ nanotubes show the lowest PL intensity, which promotes electron–hole separation and migration. Therefore, it can be reasoned that Bi₂S₃ nanotubes, with more active sites and a lower recombination rate of electron–hole pairs, demonstrates higher photocatalytic performance. To illustrate the photostability, the XRD pattern and SEM images of Bi₂S₃ nanotubes after photocatalytic H₂ production were further investigated. The XRD pattern of Bi₂S₃ nanotubes after photocatalytic H₂ production is presented in Fig. 6e. The morphology after photocatalytic H₂ production is also confirmed by SEM images. These results prove that Bi₂S₃ nanotubes exhibit good stability after photocatalytic H₂ production.

Fig. 6

a Photocatalytic H₂ evolution, b H₂ yield and c recycling test of the Bi₂S₃ nanotubes, nanoflowers and nanorods in a mixed aqueous solution containing Na₂SO₃ and Na₂S under simulated sunlight irradiation, d Four cycles of H₂ yield for the Bi₂S₃ nanotubes, e the XRD pattern and f SEM images of Bi₂S₃ nanotubes after photocatalytic H₂ production

Conclusion

In summary, Bi₂S₃ samples with three morphologies are successfully designed for photocatalytic reactions by a facile hydrothermal method at 120 °C by changing the sulfur source. Compared with the Bi₂S₃ nanoflowers and nanorods, the Bi₂S₃ nanotubes exhibit the best hydrogen production performance and good stability. The above characterization analysis confirms that the improvement in performance of the Bi₂S₃ nanotubes possibly originates from the combined effect of their larger pore structure and wider reactive channels both inside and outside the Bi₂S₃ nanotube surface. This research may provide insight for the exploitation of photocatalytic hydrogen production and extend its application to other energy fields.