Fabrication of Two-Dimensional B-Doped C₃N₄ Nanosheet-Encapsulated One-Dimensional Rod-like Mo-MOF-Derived MoS₂ Heterojunctions for Enhanced Photocatalytic Ethanol Conversion and Synergistic Hydrogen Production

Abstract

The photocatalytic conversion of ethanol and the simultaneous development of hydrogen technology play a role in solving the energy crisis and reducing environmental pollution. In this research, rod-like M-MoS₂ serves as a channel for charge transfer, leading to superior photocatalytic activity compared to H-MoS₂. Further, two-dimensional (2D) B-doped C₃N₄ (BCN) nanosheets were anchored on the one-dimensional (1D) rod-like M-MoS₂ surface to form a 1D/2D heterojunction, with M-MoS₂/BCN-0.08 (mass ratio of M-MoS₂:BCN of 0.08:1) exhibiting the highest photocatalytic performance. Under visible light irradiation, the ethanol conversion rate reached 1.79% after 5 h of photocatalytic reaction per gram of catalyst, while generating 421 μmol of 2,3-butanediol (2,3-BDO), 5460 μmol of acetaldehyde (AA), and 5410 μmol of hydrogen gas (H₂). This different characterization provides evidence that a significant amount of photoinduced electrons generated in BCN under illumination conditions rapidly transfer to the conduction band (CB) of M-MoS₂ through the rod-like structure of M-MoS₂, and finally transfer to Pt to promote the production of hydrogen gas. The photoinduced holes in the valence band (VB) of M-MoS₂ are rapidly consumed by ethanol upon transferring to BCN, effectively separating the photoinduced electron–hole pairs and resulting in superior photocatalytic performance.

1. Introduction

Ethanol, as the most basic monohydric alcohol, is used to produce value-added chemical compounds. These compounds enable a number of various substances to be obtained, relating to large-scale final products and intermediates such as commodity chemicals (hydrogen, ethylene, acetaldehyde (AA), ethyl acetate, 2,3-butanediol (2,3-BDO), etc.). The use of photocatalytic technology to convert abundant monohydric alcohols into high value-added chemicals, while generating a large amount of hydrogen gas, exhibits significant potential for solving the energy and environmental pollution crises [1]. At present, the photocatalysts for converting ethanol are mainly TiO₂, ZnS, and CdS [2,3,4]; however, these photocatalysts have the disadvantages of poor light absorption and low long-term stability. Thus, it is an urgent priority to develop new photocatalysts for photocatalytic ethanol conversion.

Due to its good visible light response, excellent thermal stability, and low cost, g-C₃N₄ has shown significant advantages in the field of photocatalysis [5,6,7,8]. For instance, carbon nitrides, including non-graphitic carbon nitrides, demonstrated high catalytic activity in the Knoevenagel reaction, synthesis of dimethyl carbonate via transesterification of ethylene carbonate, Friedel–Crafts acylation, oxidation of cyclic olefin intepoxides, etc. [9,10,11]. However, the inferior photogenerated electron–hole separation efficiency and limited active sites lead to the extremely unsatisfactory photocatalytic efficiency of the bare g-C₃N₄ [12,13,14]. Therefore, a series of modification methods, such as dimensionality control, doping, and heterostructure construction with other materials, have been implemented to boost the photocatalytic performance of g-C₃N₄ [15,16,17].

Metal–organic frameworks (MOFs), as a species of porous crystals with specific topological structures, are being widely used in gas adsorption, catalysis, sensing, and other fields [18,19]. Semiconductor materials, with nanorods, nanosheets, and hollow structures synthesized using MOFs as templates, can effectively improve the charge transfer efficiency and absorption properties in photocatalytic processes [20,21]. Therefore, metal sulfides derived from MOFs have unique advantages over traditional sulfides and are used to produce composites with g-C₃N₄, significantly facilitating charge transfer and enhancing the photocatalytic efficiency [22,23,24]. For example, Fan et al. derived a new hollow, cubic CoS structure from Co-ZIF-9 and successfully constructed S-C bonds between CoS and g-C₃N₄, achieving efficient hydrogen evolution [25]. Qi et al. designed C-doped Ni₃S₄/Ni₂P hybrid co-catalysts decorated on g-C₃N₄ using Ni-MOF as a template, exhibiting a 10 times higher photocatalytic performance compared to the pristine g-C₃N₄ [26]. Hu et al. prepared a three-component C-Cu₂-xS@ g-C₃N₄ heterojunction using HKUST-1 as a precursor and observed 23-fold enhanced activity in comparison to C₃N₄ [27].

Most recently, molybdenum disulfide (MoS₂) has been widely employed for constructing heterojunctions with C₃N₄ due to its good electrical conductivity, broad light absorption range, and appropriate band gap [28,29]. For example, Nagaraja et al. synthesized MoS₂ microflowers and g-C₃N₄ heterostructures using a one-pot solvothermal method, exhibiting significantly enhanced photocatalytic activity [30]. Li et al. successfully constructed a g-C₃N₄/Cu/MoS₂ Z-type heterojunction with excellent photocatalytic activity employing the two-step calcination method [31]. As for the MOF-derived MoS₂, an MoS₂/CdS photocatalyst was constructed, and the Mo-MOF-derived MoS₂ provided a sustainable cocatalyst for CdS [32]. In addition, Bi₄O₅Br₂ was anchored onto the rod-like MOF-derived MoS₂ to form a p-n heterojunction, which exhibited a distinctly improved photocatalytic performance [33]. However, the use of Mo-MOF-derived MoS₂ and g-C₃N₄ to construct heterostructures for use as photocatalysts for efficient ethanol conversion has not been extensively studied.

In a previous work, a B-doped C₃N₄ nanosheet (BCN) was first synthesized and further combined with ZIF67, which exhibited an excellent photocatalytic performance [34]. Here, rod-shaped MoS₂ was synthesized using a Mo-MOF rod-like structure as a template and 1D/2D heterostructures were further constructed with B-doped g-C₃N₄ nanosheets. The unique structure of the BCN nanosheets encapsulating the rod-like MoS₂ resulted in an enhanced light absorption capability, displaying outstanding photocatalytic behavior. Moreover, a plausible reaction mechanism and a charge transfer pathway were also proposed according to the characterization information.

2. Results and Discussion

Figure 1a–c show the SEM images of H-MoS₂, Mo-MOF, and M-MoS₂, respectively. The hydrothermally synthesized MoS₂ is an assembly of numerous nanosheets with severe aggregation. Furthermore, the synthesized Mo-MOF exhibits a rod-like structure with an interlocked arrangement and a smooth surface, indicating good dispersibility. Moreover, the M-MoS₂ prepared using Mo-MOF as a template maintains the rod-like structure of Mo-MOF, with an interlocked arrangement of rod-like structures, demonstrating that the M-MoS₂ inherits the excellent dispersibility from Mo-MOF. In order to investigate the effect of the changes in morphology on material properties, the photocatalytic hydrogen evolution performance of H-MoS₂ and M-MoS₂ was tested under the irradiation of a 300 W xenon lamp in a mixture of 45 mL ethanol and 5 mL aqueous solution (Figure 2a,b). The photocatalytic hydrogen evolution rate of the H-MoS₂ synthesized by the conventional hydrothermal method is only 7.3 μmol g⁻¹ h⁻¹, while that of the M-MoS₂ derived from Mo-MOF reaches 51.8 μmol g⁻¹ h⁻¹. Compared to H-MoS₂, the unique rod-like structure of M-MoS₂ serves as a channel for charge transfer, promoting the transport of photogenerated electron–hole pairs and thus resulting in an excellent photocatalytic activity. Figure 3a–d show the SEM images of the 1D/2D M-MoS₂/BCN-X composites. It can be observed that all four synthesized composite materials exhibit a well-coated structure of BCN on M-MoS₂, with a well-maintained rod-like morphology of M-MoS₂. Different amounts of BCN are dispersed around the M-MoS₂ in each sample, but overall, there is no significant difference in the morphology of the four samples.

TEM characterization was performed on the M-MoS₂/BCN-0.08 to investigate the morphological and structural features, as well as the nature of the bonding, between M-MoS₂ and BCN in the composite material. As shown in Figure 4a, the rod-shaped M-MoS₂ is dispersed with a large amount of BCN around it. The further magnification in Figure 4b reveals that the BCN surrounding M-MoS₂ consists of only a few layers, appearing as translucent flakes, and the two are tightly bound together, revealing the presence of a close heterointerface between M-MoS₂ and BCN. Figure 4c shows a high-resolution TEM image of the M-MoS₂/BCN-0.08, with a measured lattice spacing of 0.326 nm, corresponding to the (002) plane of BCN [35]. However, the M-MoS₂ has no apparent lattice spacing, which indicates the formation of an amorphous structure. Figure 4d displays a further STEM mapping of the examined samples, where the B and N elements are detected around the Mo and S elements, indicating that a large amount of thin and translucent BCN is anchored to the surface of the M-MoS₂. Thus, a close contact heterointerface exists between M-MoS₂ and BCN.

Figure 5a presents the XRD pattern of M-MoS₂. The relatively poor crystallinity of the synthesized M-MoS₂ provides evidence that M-MoS₂ is amorphous, which is consistent with the HRTEM observation [36]. In addition, diffraction peaks corresponding to the (100) and (002) crystal planes of BCN can be observed in the entire pattern [37,38]. Figure 5b shows the XRD patterns of M-MoS₂ and M-MoS₂/BCN-X. Due to the weak crystallinity of M-MoS₂, only the diffraction peaks of BCN are clearly observable.

Figure 6a displays the total XPS spectrum of the M-MoS₂/BCN-0.08 composite material. The existence of the C, N, B, and Mo elements can be clearly identified. The further analysis of the B 1s fine spectrum in Figure 6b shows a binding energy peak at 191.7 eV, which corresponds to the typical N-B-N bond, confirming the existence of the BCN in the composite material [39]. Figure 6c presents the fine spectrum of Mo 3d, and two peaks at 229.3 eV and 232.3 eV corresponding to Mo 3d 5/2 and Mo 3d 3/2, respectively, can be confirmed, suggesting the presence of Mo⁴⁺ [40]. A weak peak at 235.8 eV is also observed, suggesting that a small amount of Mo⁴⁺ was slightly oxidized to Mo⁶⁺ during the material preparation process. Figure 6d shows the fine spectrum of S 2p, with two characteristic peaks at 162 eV and 163.2 eV, confirming the presence of S²⁻ [41]. Through SEM, TEM, XRD, and XPS analyses, the successful preparation of M-MoS₂ and M-MoS₂/BCN-X was confirmed. M-MoS₂ exhibits a unique rod-like structure, different from that produced via conventional hydrothermal methods, and demonstrates an excellent photocatalytic performance. M-MoS₂/BCN-X, on the other hand, exhibits a structure where the nanosheets envelop the rod-like structure, with a well-defined heterojunction at the interface between the two components.

The main products, H₂, AA, and 2,3-BDO, were determined through chromatographic analysis. The results showed that the prepared photocatalyst achieved the conversion of ethanol while generating a large amount of hydrogen gas. The conversion of ethanol per gram of catalyst within 5 h was calculated by measuring the amount of ethanol before and after the reaction using chromatography, as depicted in Figure 7a. The ethanol conversion of the BCN was 0.86%; after being combined with M-MoS₂, the photocatalytic performance of the materials was improved to some extent. M-MoS₂/BCN-0.08 exhibited the highest photocatalytic performance: the ethanol conversion reached 1.79%. This indicates that the material obtained through the combination of the M-MoS₂ and BCN had a remarkably enhanced photocatalytic performance. Figure 7b presents the production rates of each substance per gram of photocatalyst per hour. It can be observed that the BCN exhibits weak photocatalytic activity, producing 534 μmol of hydrogen gas, 586 μmol of AA, and 43.9 μmol of 2,3-BDO per gram of BCN per hour under xenon lamp irradiation. When combined with M-MoS₂, the resulting composite material exhibits a significantly enhanced photocatalytic performance, particularly M-MoS₂/BCN-0.08 which generates 1082 μmol of hydrogen gas, 1092 μmol of AA, and 84.2 μmol of 2,3-BDO under xenon lamp irradiation. Based on the ethanol conversion rate and the yield of AA and 2,3-BDO within 10 h, the selectivity of the different photocatalysts for AA and 2,3-BDO after 10 h of reaction can be calculated using a carbon content balance (Equations (1)–(3)), as follows:

$${\mathrm{S}}_{\mathrm{A}\mathrm{A}}={2\mathrm{n}}_{\left(\mathrm{A}\mathrm{A}\right)}/{2\mathrm{n}}_{\left(\mathrm{e}\right)}$$

(1)

$${\mathrm{S}}_{\mathrm{2,3}-\mathrm{B}\mathrm{D}\mathrm{O}}={4\mathrm{n}}_{(\mathrm{2,3}-\mathrm{B}\mathrm{D}\mathrm{O})}/{2\mathrm{n}}_{\left(\mathrm{e}\right)}$$

(2)

$${\mathrm{S}}_{\mathrm{o}\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{r}}=({2\mathrm{n}}_{\left(\mathrm{e}\right)}-{2\mathrm{n}}_{\left(\mathrm{A}\mathrm{A}\right)}-{4\mathrm{n}}_{\left(\mathrm{2,3}-\mathrm{B}\mathrm{D}\mathrm{O}\right)})/{2\mathrm{n}}_{\left(\mathrm{e}\right)}$$

(3)

where n(AA) represents the amount of substance that produces AA within 10 h; n(2,3-BDO) represents the amount of substance that produces 2,3-BDO within 10 h; n(e) represents the amount of substance that converts ethanol within 10 h; and S represents the selectivity of the corresponding product. As shown in Figure 7c, the results indicate that the addition of M-MoS₂ does not affect the selectivity of the products obtained after the ethanol conversion. The reason for this phenomenon is that the loading of M-MoS₂ does not affect the desorption of the hydroxyethyl radical (•CH(OH)CH₃). Therefore, compared with the pure BCN, the selectivity of the products formed by ethanol conversion remains unchanged.

To investigate the effect of the water content in solvents on the ethanol conversion and hydrogen production performance, the M-MoS₂/BCN-0.08 photocatalysts were used for photocatalytic experiments in ethanol solutions with different water content. As shown in Figure 8a, it can be seen that when the water content is 0%, almost no hydrogen gas is generated and the substances generated by the conversion of ethanol are also extremely low. When a trace amount of aqueous solution is added to the solvent (5%), its photocatalytic performance is significantly enhanced, especially with a sharp increase in the generated hydrogen gas. The significant enhancement in the photocatalytic performance can be attributed to the fact that the addition of aqueous solutions consumes a large number of electrons, thereby promoting the production of hydrogen gas, and there are more holes that can oxidize ethanol to generate a large amount of •CH(OH)CH₃. Moreover, the reduction potential of water is higher than that of 2,3-BDO and AA, so the addition of water can also effectively inhibit the reverse reaction of 2,3-BDO and AA. When the water content increases to 10%, the photocatalytic performance is enhanced and the selectivity of the products formed by ethanol conversion changes (Figure 8b). This is because the •CH(OH)CH₃ generated by the hole oxidation of ethanol can interact with water to form hydrogen bonds, causing the partial desorption of •CH(OH)CH₃ from the surface of the photocatalyst, and thereby promoting C-C bond coupling and generating a significant amount of 2,3-BDO. Therefore, the selectivity of 2,3-BDO is significantly improved. When the water content increases to 20%, the production of the hydrogen increases, while the conversion rate of the ethanol decreases. The reason for this phenomenon may be that the photocatalysts compete for the adsorption of water and ethanol. In Figure 8c, the electron paramagnetic resonance (EPR) spectrum shows a significant amount of •CH(OH)CH₃ in the solution under light conditions for the generation of AA and 2,3-BDO. This indicates that the photocatalysts can convert a large amount of ethanol under illumination.

In order to evaluate the stability of the photocatalysts, three cycles of photocatalytic conversion of ethanol were conducted on M-MoS₂/BCN-0.08, the photocatalyst with the best photocatalytic performance, with each cycle lasting for 5 h. As shown in Figure 9a, with the increase in the photocatalytic time and the number of cycles, the photocatalytic M-MoS₂/BCN-0.08 still maintains good stability. The selectivity of the products formed by the ethanol conversion also remains stable (Figure 9b). The SEM and XRD tests on the photocatalyst after multiple cyclic tests are shown in Figure 9c and Figure 9d, respectively. It can be seen that the morphology and structure of the photocatalyst basically remained unchanged, indicating that the catalyst has good stability under the illumination conditions.

UV-Vis diffuse reflectance spectra were measured to investigate the optical absorption properties of the samples. From Figure 10a, it can be observed that the M-MoS₂ derived from Mo-MOF exhibits moderate light absorption overall but shows a certain absorption in the visible and near-infrared range. In addition, the BCN exhibits a strong absorption ability in the UV region, with the absorption edge located at approximately 470 nm. Compared to the BCN, M-MoS₂/BCN-0.08 shows a slight red shift in the absorption edge in the visible region, indicating an enhanced absorption ability in the visible range and a change in the bandgap. The bandgap widths of the materials were calculated using the Kubelka–Munk equation [42]. As shown in Figure 10b, the bandgap widths of the BCN, M-MoS₂, and M-MoS₂/BCN-0.08 are determined to be 2.62 eV, 2.32 eV, and 2.52 eV, respectively. The composite photocatalyst M-MoS₂/BCN-0.08 has a smaller bandgap width compared to the BCN, indicating that M-MoS₂ enhances the visible light absorption ability of the BCN and the narrower bandgap promotes the charge transfer capability and improves the photocatalytic activity of the material. Figure 10c displays the Mott–Schottky slopes of the BCN and M-MoS₂, which are both positive, indicating that the two materials are typical n-type semiconductors. Further, the flat band potentials of the BCN and M-MoS₂ are determined to be −0.9 V and −0.39 V, respectively, vs. the SCE based on the intercept results of the graph. Since the flat band potential of the n-type semiconductors is usually higher by about 0.2 V compared to the conduction band (CB) potential, the CB potentials of the BCN and M-MoS2 are calculated to be −1.1 V and −0.59 V (−0.86 and −0.35 V vs. NHE), respectively. Meanwhile, the VB potentials of the BCN and M-MoS₂ are calculated to be 1.76 V and 1.97 V, respectively, using Eg = E_VB − E_CB.

To detect the electron transfer pathways in the composites, EPR measurements were carried out by dispersing the samples in a methanol solution containing a DMPO spin trapping agent. According to the concentration of the captured superoxide radicals in the solution, the electron concentration of the CB of the materials can be inferred, thus providing specific information on the electron transfer in the photocatalyst. As shown in Figure 10d, no significant DMPO-•O₂⁻ signals are observed for all three samples under the dark environment, while the DMPO-•O₂⁻ signals for all samples are significantly enhanced under the light conditions. This suggests that the electrons generated from the three groups of samples can reduce O₂ to O₂⁻ in the solution under light irradiation. Among them, the M-MoS₂/BCN-0.08 shows the strongest DMPO-•O₂⁻ signal, which illustrates that this composite generates the most electrons for O₂ reduction under the same conditions. Therefore, the electrons on the CB of BCN continuously transfer to the CB of M-MoS₂ during the photocatalytic process, providing evidence that a type II heterojunction between M-MoS₂ and BCN is formed. Figure 10e displays the band structure of the M-MoS₂ and BCN, illustrating the specific pathways for the electron–hole transfer during the photocatalytic process. The type II heterojunction constructed by these two materials effectively improves the separation of the photo-generated charge carriers, leading to an enhanced photocatalytic performance. To further investigate the recombination behavior of the photo-generated charge carriers, photoluminescence (PL) spectroscopy was employed to evaluate the lifetime of the charge carriers in different samples, as shown in Figure 10f. The M-MoS₂ exhibited a strong PL emission peak at around 600 nm, while BCN showed a reduced emission peak compared to M-MoS₂, indicating the shorter lifetime of the photo-generated charge carriers in both materials. For all composites of M-MoS₂ with BCN with different ratios, a significant decrease in the PL intensity was observed, indicating that the heterojunction interface between M-MoS₂ and BCN effectively accelerates the separation of the electron–hole pairs. Among all the composite materials, M-MoS₂/BCN-0.08 exhibited the lowest PL intensity, indicating the longest lifetime of the photo-generated charge carriers, which corresponds to its excellent photocatalytic hydrogen evolution performance. The PL spectra suggest that the unique nanosheet-coated rod-like structure and the type II heterojunction formed by M-MoS₂ and BCN can effectively spatially separate the photo-generated electron–hole pairs, thereby enhancing the lifetime of the charge carriers in the material.

Photoelectrochemical tests were conducted on M-MoS₂, BCN, and M-MoS₂/BCN-X to analyze the separation and transfer abilities of the photo-generated electron–hole pairs. As shown in Figure 11a, M-MoS₂ and BCN exhibit low PC responses. However, the composites formed by the two show a significantly improved PC response, indicating that the heterojunction structure constructed by M-MoS₂ and BCN effectively enhances the separation ability of the photo-generated charge carriers. The highest PC response for M-MoS₂/BCN-0.08 also corresponds to its optimal photocatalytic hydrogen evolution performance. The impedance test results of the materials are shown in Figure 11b, where a bigger curvature radius in the impedance spectrum indicates a faster charge transfer resistance. Among them, the curve with the smallest curvature radius corresponds to M-MoS₂/BCN-0.08, indicating that this material has the best separation and transfer ability of the photo-generated carriers. The LSV curves are shown in Figure 11c. Compared with the M-MoS₂ and BCN, M-MoS₂/BCN-X shows varying degrees of increased cathodic currents. The maximum cathodic current of the M-MoS₂/BCN-0.08 may be caused by the rapid electron migration between the contact interface, implying that the heterojunction prompts the electron transfer and enhances the photocatalytic activity. The improvement in the performance is related to the rapid charge transfer at the material interface and the significant increase in the lifetime of the photo-generated charge carriers after the formation of the composite material. The rod-like structure of Mo-MoS₂ can serve as a special channel for the charge transfer, resulting in a faster charge transfer rate after the transfer of electrons from BCN CB to M-MoS₂ CB. The optimal electrochemical performance of the M-MoS₂/BCN-0.08 is consistent with its best photocatalytic hydrogen evolution performance.

The proposed charge transfer mechanism of the 1D/2D M-MoS₂/BCN-0.08 heterojunction under illumination conditions is presented in Figure 12. In comparison to MoS₂ prepared using the conventional hydrothermal method, M-MoS₂ exhibits a rod-like structure similar to Mo-MOF, with a large specific surface area and pore volume due to its spatial dispersion. This results in the exposure of more active sites, leading to superior photocatalytic performance. Additionally, the unique rod-like structure of M-MoS₂ serves as a channel for charge transfer, facilitating faster electron transfer during the process. Moreover, M-MoS₂ has a significant absorption ability throughout the visible light region, and the composite with BCN causes a redshift in the material, reducing the energy required for electron excitation in BCN. Furthermore, the reduced bandgap of the M-MoS₂/BCN-0.08 enhances its visible light absorption capacity.

The charge transfer mechanism of the 1D/2D M-MoS₂/BCN-0.08 heterojunction is further analyzed. Under illumination conditions, both the M-MoS₂ and BCN generate photoinduced electron–hole pairs. Excited-state electrons migrate to the CB of M-MoS₂ and BCN, leaving photoinduced holes in the corresponding VB. Due to the potential difference between the CB of the BCN and M-MoS₂, the electrons in the CB of BCN spontaneously transfer to the CB of the M-MoS₂. With the continuous progress of the reaction, the accumulated electrons on the CB of M-MoS₂ react gradually with the adsorbed H⁺ to form atomic hydrogen. Then, atomic hydrogen migrates to Pt particles to form H₂. During this process, ethanol acts to consume the photoinduced holes generated in the photocatalytic reaction system. The ethanol molecules adsorbed on the catalyst surface are first oxidized to •CH(OH)CH₃, and then some of the •CH(OH)CH₃ is desorbed onto the catalyst surface for C-C bond coupling to form 2,3-BDO. Some of the •CH(OH)CH₃ that fails to be desorbed in a timely manner will be further oxidized to AA by more holes and finally will be oxidized to other substances such as acetic acid or CO₂.

Compared to the normal hydrogen electrode (NHE), the CB potentials of M-MoS₂ and BCN are −0.35 V and −0.9 V, respectively, indicating that both the M-MoS₂ and BCN have sufficiently high levels of CB energy to drive the photocatalytic production of H₂. The type II heterojunction formed by M-MoS₂ and BCN effectively utilizes the photoinduced electrons generated by both materials. Moreover, the thin nanosheets of BCN and the unique rod-like structure of M-MoS₂ allow for the uniform coverage of BCN on the surface of M-MoS₂. During the photocatalytic hydrogen evolution process, the large amount of photoinduced electrons generated in BCN under illumination conditions rapidly transfer to the CB of M-MoS₂ through the rod-like structure of M-MoS₂, while the photoinduced holes in the VB of M-MoS₂ are rapidly consumed by ethanol upon transferring to BCN and promote the generation of AA and 2,3-BDO, effectively separating the photoinduced electron–hole pairs and resulting in superior photocatalytic performance.

3. Materials and Methods

3.1. Synthesis of B-Doped g-C₃N₄ Nanosheets

The B-doped g-C₃N₄ nanosheets were prepared using a thermal polymerization and exfoliation method [43,44]. Specifically, 5 g of melamine (C₃H₆N₆) and 40 mg of boric acid (H₃BO₃) were mixed and ground. Then, the collected powders were transferred to a covered alumina crucible, heated from room temperature to 550 °C at a rate of 5 °C/min in a muffle furnace, and soaked for 4 h. The obtained solid was fully ground to obtain a B-doped g-C₃N₄ nanosheet powder, which was then placed in an open alumina crucible and heated again under the condition of 500 °C for 2 h in a muffle furnace at a rate of 5 °C/min, resulting in B-doped g-C₃N₄ nanosheets.

3.2. Preparation of MoS₂ Nanosheets (H-MoS₂)

A total of 90 mg of thiourea (C₂H₅NS) and 50 mg of sodium molybdate dihydrate (Na₂MoO₄•2H₂O) were mixed in 40 mL of deionized water, and the transparent solution was transferred to a 100 mL Teflon-lined autoclave and heated at 220 °C for 24 h. The black powder was collected, washed with deionized water three times, and dried in vacuum at 80 °C for 12 h to obtain H-MoS₂.

3.3. Preparation of Rod-like Mo-MOF and Derived MoS₂ (M-MoS₂)

Then, 3.5 g of molybdenum trioxide (MoO₃) and 1.66 g of imidazole (C₃H₄N₂) were dissolved in 250 mL of deionized water, and the mixture was refluxed at 100 °C for 12 h. After that, the precipitate was collected by centrifugation. After washing three times with deionized water, Mo-MOF microrods were obtained by vacuum drying at 70 °C. M-MoS₂ was prepared using Mo-MOF as a template. Specifically, 0.2 g of the prepared Mo-MOF and 1.6 g of thiourea (CH₄N₂S) were placed in separate open alumina crucibles, and the crucibles were placed downstream and upstream of a tubular furnace. After purging with N₂ for 30 min to remove the oxygen from the furnace, the temperature was increased to 600 °C at a rate of 5 °C/min under an N₂ atmosphere and maintained for 3 h to obtain the black M-MoS₂.

3.4. Preparation of 1D/2D M-MoS₂/BCN Heterostructure

The 1D/2D M-MoS₂/BCN heterostructure was prepared using a mechanical grinding method that enables precise control of the material ratio. Specifically, 0.2 g of the prepared BCN and a certain amount of M-MoS₂ were placed in a quartz grinding dish and thoroughly ground for 30 min to obtain M-MoS₂/BCN-X, where X represents the mass ratio of M-MoS₂ to BCN. Four different mass ratios of the composite materials were prepared, namely M-MoS₂/BCN-0.02, M-MoS₂/BCN-0.05, M-MoS₂/BCN-0.08, and M-MoS₂/BCN-0.11.

3.5. Characterization of the Catalyst

The morphologies of the samples were analyzed via field emission scanning electron microscopy (FE-SEM, TESCAN MIRA3 LMH). Transmission electron microscopy (TEM) images were characterized using a transmission electron microscope (TEM, JEOL 2100F, JEOL, Tokyo, Japan). The phase and structure of the sample were measured via X-ray diffraction (XRD; Panalytical Aeries Research) with Cu Ka radiation. The elemental composition and valence states were determined via X-ray photoelectron spectroscopy (XPS, Thermo Scientific ESCALAB 250Xi, Waltham, MA, USA). The light absorption was evaluated using PerkinElmer Lambda 1050+ UV-V (PerkinElmer, Waltham, MA, USA) reflectance spectroscopy with BaSO₄ as the standard reference material. The photoluminescence (PL) measurement was conducted on the FLS980 fluorometer. Electron paramagnetic resonance (EPR) was carried out by dispersing the samples in a methanol solution containing a DMPO spin trapping agent, using a Brucker EPR A 200 W spectrometer (Brucker, Billerica, MA, USA). The photoelectrochemical measurements, including transient photocurrent curves (I–t curve, photocurrent vs. SCE at 0.598 V), electrochemical impedance spectroscopy (EIS, at 0.2 V potential vs. SCE), and linear sweep voltammogram (−0.6 V to 0 V vs. SCE), were performed on an electrochemical workstation (Metrohm Autolab; Nova, Waltham, MA, USA) with a self-made standard three electrode battery. A conductive glass electrode coated with the samples, a Pt electrode, and a saturated calomel electrode (SCE) were used as the working electrode, counter electrode, and reference electrode, respectively. To prepare the fluorine-doped tin oxide (FTO) working electrode, 5 mg of the sample was dispersed in a mixed solution consisting of 50 μL Nafion and 500 μL ethanol using ultrasound to obtain a slurry. Afterwards, the slurry was pre-immersed into the FTO glass protected by paper tape at the boundary. After air drying, the paper tape was exposed to obtain the working electrode. All electrochemical measurements were performed using 0.2 M Na₂SO₄ aqueous solution as the electrolyte and a 300 W Xe lamp as the light source.

3.6. Photocatalytic Hydrogen Production

A total of 50 mg of the photocatalyst with 1 wt.% Pt (H₂PtCl₆•6H₂O) added as a cocatalyst was dispersed in 50 mL of ethanol–water solution with varying water contents (0%, 5%, 10%, and 20%). Prior to the reaction, the reactor was purged with Ar gas to remove air. During the reaction, continuous stirring was maintained to prevent the aggregation of the photocatalyst under 300 W Xe lamp irradiation. The production of hydrogen, 2.3-butanediol, and acetaldehyde was measured using the Agilent GC-950 gas chromatograph (Agilent, Santa Clara, CA, USA). During the stability experiment, the liquid after 5 h of photocatalytic reaction was centrifuged at 10,000 rpm for 10 min at high speed. The photocatalyst was washed with ethanol to remove impurities and then dried at 60 °C for 12 h. The photocatalytic cycling experiment was continued using the dried photocatalyst. The stability test was conducted over three cycles, with each cycle lasting for 5 h.

4. Conclusions

M-MoS₂ was derived using Mo-MOF as the template, and the M-MoS₂/BCN-X composite photocatalyst was prepared via mechanical grinding with BCN. M-MoS₂ derived from Mo-MOF has a rod-like structure and a photocatalytic hydrogen evolution rate of 51.8 μmol g⁻¹ h⁻¹, which is 7.1 times that of the H-MoS₂ synthesized using the traditional hydrothermal method. This is due to its special rod-like structure that can also act as a fast channel for charge transport. A photocatalyst with excellent photocatalytic performance was prepared by combining the 1D rod-like structure of M-MoS₂ and 2D BCN. In an ethanol aqueous solution with a water content of 10%, 1082 μmol of hydrogen gas, 1092 μmol of AA, and 84.2 μmol of 2,3-BDO were produced per gram of M-MoS₂/BCN-0.08 per hour. The special one-dimensional rod-like structure of M-MoS₂ provides an effective composite area for the loading of two-dimensional BCN. The electrons in the CB of BCN are transferred to the CB of M-MoS₂, and the rod-like structure as a transmission channel accelerates the electron transfer. The holes located in the VB of M-MoS₂ are uniformly transferred to the BCN, which promotes the separation of the photogenerated carriers and leads to a type II heterojunction with excellent photocatalytic properties.