Bilayer MoSe₂/HfS₂ Nanocomposite as a Potential Visible-Light-Driven Z-Scheme Photocatalyst

1. Introduction

In order to set up a sustainable society, photocatalytic water splitting for hydrogen production has been deemed as an effective route to solve the problems of environmental pollution and energy shortage [1]. Owing to the groundbreaking work by Honda and Fujishima in 1972 [2], a variety of semiconductor materials have been extensively investigated to explore high-performance photocatalysts for water decomposition [3,4,5,6,7]. However, a majority of one-component photocatalysts, such as TiO₂ and ZnO, can only utilize a small amount of the solar energy and the lifetimes of photoinduced electron–hole pairs in these materials are short, which leads to the problem that the photocatalytic efficiency is low and hampers their future applications [8,9]. For the sake of overcoming these disadvantages, many researchers have found that the construction of heterostructures, which are composed of different materials and can supply much greater control of the electronic and optical properties, is able to effectively enhance the catalytic activity [10,11,12,13,14,15].

In particular, great expectations have been placed on the Z-scheme photocatalytic mechanism—which includes two-step excitation and is illuminated by natural photosynthesis in plants [16]—for improving the utilization efficiency of sunlight [17,18,19,20,21,22,23,24]. Generally speaking, the Z-scheme photocatalytic system is made up of three parts: Catalysts for the hydrogen evolution reaction (HER), oxygen evolution reaction (OER), and the redox mediator for carrier migration [25]. According to this mechanism, although the isolated components cannot accomplish overall water splitting, the combined systems can decompose water into hydrogen and oxygen, which can broaden the scope of the promising photocatalysts. Moreover, the Z-scheme photocatalytic composites possess strong redox abilities and can benefit the separation of photogenerated carriers because of the occurrence of HER and OER on different layers [26]. Nevertheless, the redox mediators in Z-scheme photocatalysts may induce undesired back reactions and seriously affect the photocatalytic performance [27]. Z-scheme systems without mediators, which are named as direct Z-scheme photocatalysts, can avoid the problems caused by mediators and are easier to experimentally fabricate owing to their simpler structures. Therefore, the direct Z-scheme systems have been extensively studied [28,29,30,31,32].

Recently, various two-dimensional (2D) materials, such as graphene [33], g-C₃N₄ [34], MoSe₂ [35], and HfS₂ [36] have been predicated and fabricated [37]. Owing to the maximized specific surface area, high charge migration, and unique electronic properties derived from the quantum confinement effect, 2D materials and related heterostructures have attracted much attention as high-performance photocatalysts for water [38,39,40]. However, being limited by the band edge positions, the isolated MoSe₂ and HfS₂ monolayers can only be used for the HER and OER [41,42], respectively. Fortunately, the band edge positions of MoSe₂ and HfS₂ single layers present staggered-type (type II) band alignment characteristics [43,44], which are beneficial for establishing the Z-scheme composites. Therefore, we have designed MoSe₂/HfS₂ bilayer nanocomposites and studied their geometric structures, band structures, density of states, charge transfer, stress effect, and optical properties by employing the hybrid density functional theory.

2. Computational Method

In this work, all of the density functional theory calculations were employed with the Vienna ab initio simulation package (VASP) [45]. The frozen-core projector augmented wave (PAW) was used to describe the interaction between the core and valence electrons [46,47]. The generalized gradient approximation (GGA) [48] of the Perdew–Burke–Ernzerhof (PBE) exchange correlation functional [49] was adopted. The charge redistribution in the MoSe₂/HfS₂ heterojunction was considered by calculating the dipole correction [50]. For the sake of avoiding the undervaluation of bandgaps calculated by the PBE method, the electronic and optical properties of these referred materials were calculated by the Heyd–Scuseria–Ernzerhof (HSE) hybrid functional by adding a part of the exact exchange interaction [51]. An exact exchange contribution of 0.25 was used in this study. The long-range van der Waals (vdW) interaction of heterojunctions was described by the Grimmes DFT-D3 method [52]. A Γ-centered 7 × 7 × 1 K-point was used to sample the 2D Brillouin Zone [53]. The cutoff energy was set as 500 eV. The convergence criteria were less than 10⁻⁵ eV for total energy and 0.01 eV/Å for Hellman–Feynman force on each atom, respectively. A vacuum space of 20 Å was inserted perpendicular to the layers to separate the neighboring slabs of heterojunctions. The band edge positions of 2D materials were calculated by subtracting vacuum levels, which were obtained by averaging the values in the LOCPOT file; they were then applied for measuring the absolute positions of energy bands.

3. Results and Discussion

3.1. Structural Stability

In this study, we chose the hexagonal MoSe₂ and HfS₂ monolayers, which are able to fit each other well. After the structural optimization, the lattice constants of MoSe₂ and HfS₂ single layers were 3.30 Å and 3.61 Å, respectively, which agree with the previous values [54]. Due to the small lattice mismatch of these single layers, we designed a heterojunction composed of 2 × 2 MoSe₂ and HfS₂ supercells. In order to probe into the steadiest stacked pattern, we studied various configurations. The steadiest configuration of the MoSe₂/HfS₂ heterojunction, whose equilibrium structure is presented in Figure 1, will be studied in the subsequent parts.

In addition, the interlayer spacing of the MoSe₂/HfS₂ heterobilayer was 3.05 Å, as displayed in Figure 1, which refers to the vertical distance from the nearest S atoms of the HfS₂ layer to the Se atoms of the MoSe₂ layer. This vertical separation between two layers is a typical vdW equilibrium distance, which is in accord with other 2D vdW heterostructures [11,44]. Thus, it was necessary to add the vdW corrections into our computation. The binding energy (E_b) of the MoSe₂/HfS₂ heterojunction can be defined as this equation: E_b = E_MoSe2/HfS2 − E _MoSe2 − E _HfS2, where E_MoSe2/HfS2, E _MoSe2, and E _HfS2 stand for the total energy of this heterostructure, isolated MoSe₂, and HfS₂ monolayers, respectively. After calculations, the bound energy of the steadiest configuration of the MoSe₂/HfS₂ heterojunction is –0.27 eV, which means that this composite is stable.

3.2. Electronic Properties

The band structures of related materials were calculated by employing the hybrid functional. As presented in

Table 1. Bandgaps and band edge positions of the related nanosheets.

Structure	Eg (eV)	EVBM (eV)	ECBM (eV)	Bandgap Type
MoSe2	2.02	−5.63	−3.61	Direct
HfS2	1.99	−6.93	−4.94	Indirect
MoSe2/HfS2	0.53	−5.69	−5.16	Direct
MoSe2/HfS2 with −6% strain	0.53	−5.69	−5.16	Direct
MoSe2/HfS2 with −3% strain	0.54	−5.70	−5.16	Direct
MoSe2/HfS2 with 3% strain	0.53	−5.69	−5.16	Direct
MoSe2/HfS2 with 6% strain	0.53	−5.69	−5.16	Direct

, the bandgaps of MoSe₂ and HfS₂ monolayers were 2.02 and 1.99 eV, respectively, which are consistent with the previous calculations [54] and mean that the two layers can both utilize the visible light irradiation. Moreover, the valence band maximum (VBM) and conduction band minimum (CBM) of the HfS₂ monolayer were both lower than those of MoSe₂ single layer, which indicates that MoSe₂/HfS₂ heterobilayer can form a type II heterostructure. As we all know, all of the direct Z-scheme systems have a typical type II band alignment structure [26]. Therefore, the MoSe₂/HfS₂ heterojunction can form a direct Z-scheme photocatalyst.

Meanwhile, the VBM of the HfS₂ single layer was able to stride over the standard oxidation potential for O₂/H₂O, whereas the CBM was 0.5 eV lower than the standard reduction potential for H⁺/H₂. So, the HfS₂ monolayer can only be used for OER. At the same time, the VBM of the MoSe₂ monolayer was 0.04 eV higher than the standard oxidation potential, while the CBM could straddle the standard reduction potential, indicating that the MoSe₂ single layer can only be applied for HER. Thus, by combining the MoSe₂ and HfS₂ single layers, the MoSe₂/HfS₂ nanocomposite can be used for photocatalytic overall water splitting. As depicted in Figure 2, the MoSe₂/HfS₂ heterostructure is a semiconductor with a direct bandgap of 0.53 eV, which is less than those of the individual MoSe₂ (2.02 eV) and HfS₂ (1.99 eV) single layers, indicating that this heterojunction can make the best of visible light and even enlarge the applied range of light to infrared light. As shown in the picture of the projected band structure, the VBM of this heterostructure is mainly composed of the MoSe₂ layer, while the CBM is primarily made up of the HfS₂ layer, which is in support of the separation of the photoinduced carriers.

In order to systematically study the photocatalytic ability of the MoSe₂/HfS₂ heterostructure, we employed a density of states (DOS) analysis. TDOS and PDOS represent the total and partial DOS, respectively. As displayed in Figure 3, the VBM of this heterojunction chiefly consisted of the Mo 4d and Se 4p states, which were derived from the MoSe₂ layer. However, the CBM was primarily made up of the Hf 5d and S 3p orbitals, which were rooted in the HfS₂ slab. Thus, the VBM and CBM of this heterojunction were separated into different layers, which is consistent with the previous analysis about the band structure.

The effective separation of the photo-generated electron–hole pairs is an essential factor for improving the photocatalytic activity. Thus, we applied the charge density difference and Bader charge analysis to investigate the carrier migrating processes. The charge density differences of this heterostructure are defined as this: △ρ = ρ_MoSe2/HfS2 − ρ_MoSe2 − ρ_HfS2, where ρ_MoSe2/HfS2, ρ_MoSe2, and ρ_HfS2 indicate the charge density of the MoSe₂/HfS₂ nanocomposite, freestanding MoSe₂, and HfS₂ nanosheets, respectively. The small bandgap (0.53 eV) in this heterostructure was favorable to enhancing the interlayer carrier transfer. As depicted in Figure 4, there is obvious charge accumulation and depletion near the interface. The MoSe₂ layer was apt to lose charge, while the HfS₂ slab tended to gain charge. Hence, the charge transfer between two layers brought about an internal electric field pointing from the MoSe₂ layer to the HfS₂ slab, which is in accordance with the previous report that the electric field generally points from the hydrogen evolution catalyst to the oxygen evolution catalyst [25]. In order to evaluate the quantity of charge transfer, we employed the Bader charge analysis. The S and Se atoms were liable to acquire electrons, while the Hf and Mo atoms were apt to lose electrons. As a whole, the migrated electrons from the MoSe₂ layer to the HfS₂ slab were 0.052 e, which is in keeping with the charge density differences analysis.

After the light irradiation, the electrons on the VB of MoSe₂ and HfS₂ slabs were excited to the CB of these layers. The internal electric field of the MoSe₂/HfS₂ nanocomposite was in favor of the electron migration from the CB of HfS₂ to the VB of MoSe₂. Meanwhile, the electric field hindered the electron transfer from the CB of MoSe₂ to the CB of HfS₂ and the hole transfer from the VB of HfS₂ to the VB of MoSe₂. Moreover, because the CBM of HfS₂ and VBM of MoSe₂ were close to each other (0.69 eV)—as displayed in Table 1—the photogenerated electrons in the CBM of HfS₂ layer could easily recombine with the holes in the VBM of the MoSe₂ layer. Finally, the photoinduced electrons in the CBM of MoSe₂ slab and the holes in the VBM of HfS₂ slab could both be preserved, which enabled the MoSe₂/HfS₂ nanocomposite to possess high redox ability. The schematic illustration of the Z-scheme photocatalytic mechanism for the MoSe₂/HfS₂ nanocomposite is illustrated in Figure 5. Therefore, in comparison with the MoSe₂ and HfS₂ monolayers, the MoSe₂/HfS₂ heterostructure can improve the separation efficiency of the photogenerated carriers and enhance the redox ability.

In addition, stress is inevitable in industrial production and may also originate from the mismatch of the lattices between different materials. Thus, we investigated the band edge positions of the MoSe₂/HfS₂ nanocomposite as a function of in-plain strains. By applying the strains in the range from −6% to 6%, the calculated band edge positions of the MoSe₂/HfS₂ heterostructure remained almost unchanged, as displayed in Figure 6. Thus, the band structures of MoSe₂/HfS₂ heterostructure are stable, which is different from the typical type II heterojunctions [11,15]. Therefore, the stable electronic property is in favor of future industrial applications.

3.3. Optical Properties

Optical absorption spectra are able to directly characterize the catalytic performance of photocatalysts. In this work, we employed the VASPKIT software to analyze the optical absorption properties. The optical absorption coefficient I(ω) can be acquired from the dynamical dielectric response function ε(ω), which can be expressed by this equation: $I(\omega )=\sqrt{2}\omega {[\sqrt{{\epsilon }_1{(\omega )}^2+{\epsilon }_2{(\omega )}^2}-{\epsilon }_1(\omega )]}^{\frac{1}{2}}$. Compared with those of the HfS₂ and MoSe₂ monolayers, the optical absorption intensity of the MoSe₂/HfS₂ heterostructure is strengthened, especially in the visible and infrared region, as shown in Figure 7. Because the HfS₂ single layer is an indirect semiconductor, the enhancement of the optical absorption mainly stems from the MoSe₂ monolayer, which possesses a direct bandgap. Moreover, the evident red shift in this nanocomposite is mainly derived from the transition from the S 3p state to the Se 4p state. Furthermore, there are obvious absorption peaks even at 400 and 600 nm. Therefore, the MoSe₂/HfS₂ can take full advantage of the visible light.

4. Conclusions

To summarize, we have designed a MoSe₂/HfS₂ bilayer heterostructure and investigated its electronic and optical properties according to the direct Z-scheme mechanism by employing the hybrid density functional theory. The computed results reveal that the HfS₂ monolayer and MoSe₂ single layer can only be used for OER and HER, respectively. By combining the MoSe₂ and HfS₂ monolayers, the MoSe₂/HfS₂ nanocomposite, which is a direct Z-scheme photocatalyst, can be used for photocatalytic overall water splitting. As depicted in the projected band structure, the VBM of this heterostructure is mainly composed of the MoSe₂ layer, while the CBM is primarily made up of the HfS₂ layer. By applying the charge density difference and Bader charge analyses, the charge transfer between two layers brought about a built-in electric field pointing from the MoSe₂ layer to the HfS₂ slab, which is in support of the separation of the photoinduced carriers. Moreover, the photogenerated electrons in the CBM of the MoSe₂ slab and the holes in the VBM of the HfS₂ slab can both be preserved, which enables the MoSe₂/HfS₂ nanocomposite to possess high redox ability. When strains are applied to the MoSe₂/HfS₂ heterostructure, the band structures of this heterostructure are stable, which is in favor of future industrial applications. Compared with those of HfS₂ and MoSe₂ monolayers, the optical absorption intensity of the MoSe₂/HfS₂ heterostructure is distinctly strengthened in the visible and infrared region. Therefore, the MoSe₂/HfS₂ heterojunction is a potential direct Z-scheme photocatalyst for photocatalytic overall water splitting. This work may provide an effective route for developments in clean and renewable energy.