Spontaneous generation of an atomically dispersed Mo and MoS₂ coupling catalyst via reaction induction transformation for enhancing local hydrogen concentration in hydrogenation

Abstract

The development of a synergistic coupling catalyst at the atomic scale for slurry-phase hydrogenation of vacuum residue (VR) is extremely challenging. Herein, we designed and constructed a robust coupling catalyst comprising Mo single atoms and MoS₂ (Mo SAs–MoS₂) using a reaction induction transformation strategy. The spontaneous generation of Mo SAs–MoS₂ synergistically promoted H₂ activation and enhanced local active hydrogen concentration in the hydrogenation of VR. Benefiting from the strong hydrogen concentration distribution in MoS₂ and Mo SAs, the catalyst revealed remarkable hydrogenation performance toward VR with a TOF_T of up to 0.39 s⁻¹, liquid product yield of 92 wt%, and coke content of 0.6 wt%. Theoretical calculations revealed that the synergistic action of Mo SAs–MoS₂ facilitated electron transfer between Mo species and reactants, reducing the desorption energy barriers for H₂ and anthracene and thereby promoting the hydrogenation of VR. This work provides a novel idea for the design of efficient coupling catalysts for strengthening local active hydrogen concentration in the hydrogenation of VR, and this concept can be applied to other hydrogenation systems.

---

Broader context

In this study, we designed and constructed a robust coupling catalyst comprising Mo single atoms (SAs) and MoS₂ (Mo SAs–MoS₂) using a reaction induction transformation strategy. The spontaneous generation of the Mo SAs–MoS₂ coupling catalyst synergistically promotes H₂ activation and enhances local active hydrogen concentration in the hydrogenation of vacuum residue (VR). The Mo SAs–MoS₂ catalyst demonstrates superior catalytic hydrogenation activity for the hydrocracking of VR. Benefiting from the strong hydrogen concentration distribution in MoS₂ and Mo SAs, the catalyst demonstrates superior catalytic hydrogenation activity, with a TOF_T of up to 0.39 s⁻¹ calculated for total metals, a liquid product yield of 92 wt%, and coke content of only 0.6 wt%. This work proposes a novel concept for the design and development of high-performance hydrogenation catalysts.

Introduction

With the increasing demand for light fuel oils,¹ efficiently converting and cleanly utilizing vacuum residue (VR) has become a crucial focus in the refining industry.² Slurry-phase hydrogenation technology is promising for converting heavy oil into lightweight oil.³ Efficient hydrogenation can better avoid coking of oil and improve the yield of the liquid product.^4,5 Therefore, the development of high-performance hydrogenation catalysts is the key to realizing the activation of hydrogen molecules and the hydrogenation of macromolecular free radicals in hydrocracking.

At present, the main types of catalyst used are molybdenum sulfide catalysts.^6–8 In hydrogenation, a catalyst precursor is converted into a MoS₂ active phase through in situ sulfidation.^9,10 However, the MoS₂ active phase gradually grows and aggregates during the hydrogenation reaction, leading to gradual deactivation of the catalyst.¹¹ Regarding this issue, our team previously proposed Mo-based single-atom catalysts (SACs) for enhancing the slurry-phase hydrogenation performance of VR.^12,13 Compared with nanocatalysts, SACs have more exposed active sites and specific coordination structures that can efficiently activate hydrogen molecules, resulting in their better activity and selectivity in hydrogenation reactions.^14–19 However, in complex catalytic reactions, it is often difficult to adsorb macromolecular reactants at a single active site.^20,21 However, coupling sites composed of metal single atoms (SAs) and clusters (Cs) or nanostructures can achieve synergistic catalysis for the complex hydrogenation of VR.^22–24

Herein, we designed and constructed a robust Mo SAs–MoS₂ coupling catalyst through the reaction induction transformation of Mo SAs–Mo Cs, which we prepared using an encapsulation–pyrolysis–etching strategy. The spontaneous generation of Mo SAs–MoS₂ synergistically promoted H₂ activation and enhanced local active H concentration, thus enhancing the efficient conversion of VR. It was also found that the enhancement of active H concentration avoids the continuous aggregation and marginal growth deactivation of the MoS₂ catalyst during the hydrogenation reaction. Density functional theory (DFT) calculations showed that the synergistic actions of the Mo SAs–MoS₂ coupling catalyst reduced the adsorption energy barriers for H₂ and anthracene, which significantly promoted the hydrogenation. These findings provide a novel idea for the design of efficient coupling catalysts by controlling local hydrogen species concentration for enhancing hydrogenation reactions.

Results and discussion

Synthesis and spontaneous generation of the Mo SAs–MoS₂ coupling catalyst

The Mo SAs–MoS₂ coupling catalyst was prepared by the reaction induction transformation of Mo SAs–Mo Cs, which were synthesized by an encapsulation–pyrolysis–etching strategy (Fig. 1). First, coal tar pitch (CTP) was selected as the carbon precursor because of its good oil solubility and high carbon graphitization. Next, melamine and ammonium molybdate molecules with oxygen-bridged [MoO₄] tetrahedral units were uniformly dispersed on CTP by strong π–π conjugation. Lamellar MgO was selected as the template for encapsulation, as it can fully integrate with the carbon substrate in pyrolysis. Then, Mo SAs–Mo Cs with a Mo content of 5.37 wt% were obtained by acid etching the MgO template. The Mo SAs–MoS₂ could be generated spontaneously by the in situ sulfurization reaction induction transformation of Mo SAs–Mo Cs. It should be emphasized that the S species mainly originated from heavy oil. The synthesized Mo SAs–Mo Cs exhibited a significant Brunauer–Emmett–Teller (BET) surface area of 504 m² g⁻¹ and pore size of 3 nm, as determined by N₂ physical adsorption/desorption (Fig. S1, ESI†). This large surface area and pore size are favorable for the carbon-supported capture of metal species and can improve mass transfer during the hydrogenation of VR. The 3D lamelliform-like structure with abundant pores was confirmed by scanning electron microscopy (SEM) (Fig. S2, ESI†).

Fig. 1 Schematic of the catalyst preparation process.

The X-ray diffraction (XRD) patterns of the Mo SAs–Mo Cs (Fig. S3, ESI†) showed broad peaks of 26° attributed to the amorphous carbon for the (002) plane, excluding the presence of metal nanoparticles or oxide crystals in the Mo SAs–Mo Cs. The transmission electron microcopy (TEM) images further showed the lamelliform-like carbon morphologies (Fig. S4a, ESI†). No visible metal nanoparticles (Fig. S4b, ESI†) could be discovered in the Mo SAs–Mo Cs by high-resolution TEM (HRTEM). Moreover, high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) indicated the presence of metal Mo clusters. The energy-dispersive spectroscopy (EDS) mapping confirmed the uniform distribution of Mo, O, N, and C species in the Mo SAs–Mo Cs (Fig. S4c, ESI†). Aberration-corrected HAADF-STEM (AC-HAADF-STEM) images (Fig. 2a and b) were obtained and further revealed the uniform dispersion of metal Mo clusters and Mo SAs. As shown in Fig. 2c, the sizes of the Mo clusters were about 6–7 Å, respectively. In addition, the Mo SAs–Mo Cs showed good hydrophobicity with static angles of 130.5° (Fig. 2d), which revealed their superior oil solubility.

Fig. 2 Structural evolution of Mo SAs–Mo Cs into the Mo SAs–MoS₂ catalyst. (a) and (b) AC-HAADF-STEM images, (c) cluster size distribution of areas (1) and (2) in (a), (d) corresponding contact angle of Mo SAs–Mo Cs. (e) and (f) AC-HAADF-STEM images, (g) corresponding distribution of the slab lengths of the MoS₂ nanosheets, and (h) corresponding contact angle of Mo SAs–MoS₂. MoS₂ adsorption on C and C–H substrates and (i) the variations of the Gibbs free energy, and (j) pDOS of the Mo 3d, S 2s, H 1s, C 1s orbitals, and total DOS.

Two typical peaks at 34° (100) and 59° (110) in the XRD pattern were attributed to MoS₂ in the Mo SAs–MoS₂ (Fig. S3, ESI†). HRTEM (Fig. S4d, ESI†) demonstrated that the MoS₂ crystal phase was generated during the hydrogenation reaction in VR. Also the AC-HAADF-STEM (Fig. 2e, f) images demonstrated that most Mo atoms could be embedded into the MoS₂ lattice. Combined with the characterization of the Mo SAs–Mo Cs precursor, it was confirmed that the Mo species were partly transformed from clusters into MoS₂, while the Mo SAs retained atomic dispersion characteristics (Fig. S5, ESI†).^12,13 The length statistics showed that the slab lengths of MoS₂ were mainly in the range of 4–12 Å. However, a large number of Mo SAs were distributed between the MoS₂ layers. This indicated that the Mo SAs induced the oriented growth of MoS₂ by activating H₂. Sulfurization experiments for Mo species with different gases were further used to prove the above inference. From the TEM images in H₂ (Fig. S7a–c, ESI†) and N₂ (Fig. S7d–f, ESI†) atmospheres, the MoS₂ nanosheets were denser in H₂ than in N₂, which demonstrated the induced growth of MoS₂ sheets under H₂. Moreover, the variations of the Gibbs free energies (Fig. 2i and Fig. S6, ESI†) for MoS₂ adsorption on the substrate were further studied on the C and active H-adsorbed C surfaces. It was found that MoS₂ adsorption was considerably more favorable on C–H surfaces than on the C surfaces. Compared to the C substrate, the pDOS in the C–H substrate (Fig. 2j) indicated that the covalent bonds were strongly aggregated near the Fermi level.²⁵ This active hydrogen was beneficial for the adsorption of MoS₂ on the substrate and could therefore avoid the continuous growth of MoS₂. The Mo SAs–MoS₂ catalyst also showed good hydrophobicity with a static angle of 140.9° (Fig. 2h), about 10° more than the Mo SAs–Mo Cs precursor, which further indicated that the Mo SAs–MoS₂ catalyst could be well dispersed in VR, revealing its better hydrogenation performance.

Structural evolution of Mo SAs–Mo Cs into the Mo SAs–MoS₂ catalyst

To further explore the local structural evolution upon the spontaneous generation of active sites, XPS was performed and the spectra of the Mo SAs–Mo Cs and Mo SAs–MoS₂ catalysts were compared (Fig. 3a). Because of the in situ change in the type of Mo species, the Mo 3d spectra were analyzed. The valence state of Mo species in Mo SAs–Mo Cs was mainly Mo⁶⁺, while in Mo SAs–MoS₂, Mo species with +4 and +5 valence states were generated at a much higher content than Mo⁶⁺, indicating that Mo was mainly in the +6 state, and the valences of the Mo species in MoS₂ were mainly +4 and +5. The work function of Mo SAs–Mo Cs (4.72 eV) and Mo SAs–MoS₂ (4.48 eV) were analyzed by UPS (Fig. S8, ESI†), which further inferred that the valence state of the Mo species decreased during the reaction.

Fig. 3 Local structural evolution of Mo SAs–Mo Cs into Mo SAs–MoS₂ catalysts. (a) XPS spectra, Mo K-edge in (b) XANES and (c) adsorption edge, (d) EXAFS and (e) EXAFS fitting curves. WT contour plots of (f) Mo SAs–Mo Cs, (g) Mo SAs–MoS₂, and (h) MoS₂.

The fine structure of the catalysts was further analyzed by X-ray absorption spectroscopy (XAS). Also, X-ray absorption near-edge spectroscopy (XANES) was performed as it can reflect information on the valence state and local configuration (Fig. 3b). The observed shift of Δμ₀ (edge step) demonstrated there was a change in the valence state (Fig. 3c), wherein the Mo K-edges of Mo SAs–Mo Cs and Mo SAs–MoS₂ were located between the MoO₂ and MoO₃ standard, suggesting the average valence states of the Mo species were close to +4 and +6, respectively. The valence state is related to the coordination environment of metal species. The EXAFS curves (Fig. 3d and Fig. S9, ESI†) showed peaks in the R space of 1.04 and 1.4 Å, ascribed to Mo–O coordination. The peaks at 1.93 and 2.85 Å could be assigned to the Mo–S coordination and the peak at 2.31 Å was ascribed to Mo–Mo coordination. Two effective peaks appeared at 1.1 and 2.4 Å⁻¹ in the WT-EXAFS of Mo SAs–Mo Cs (Fig. 3f), corresponding to Mo–O and Mo–Mo, respectively (Fig. S10, ESI†). In addition, peaks for Mo–O and Mo–S of Mo SAs–MoS₂ were observed at 2.0 and 2.9 Å⁻¹ (Fig. 3g and h), in contrast to the MoS₂ standard, further demonstrating the local structural evolution of Mo SAs–Mo Cs into Mo SAs–MoS₂. Moreover, the EXAFS fitting curves (Fig. 3e) showed that the Mo–O and Mo–Mo coordination numbers were about 4 and 6 in Mo SAs–Mo Cs, while the Mo–O and Mo–S coordination numbers were about 4 and 2 after the EXAFS fitting for Mo SAs–MoS₂, respectively (Table S2, ESI†). For comparison, the Mo SAs catalyst exhibited a Mo–O isolate configuration through XAS analysis (Fig. S11a and b, ESI†), wherein the EXAFS fitting results demonstrated a greater degree with a Mo–O coordination number of 3.62 (Fig. S11c, d and Table S2, ESI†).

Hydrogenation performance of VR

The catalytic hydrogenation evaluation of VR was conducted in an autoclave with H₂ at 7 MPa and 425 °C for 1 h. The pressure drop curve illustrates the H₂ consumption in the reaction system within 1 h (Fig. S12, ESI†). The Mo SAs–MoS₂ catalyst produced the maximum pressure drop, indicating the maximum hydrogen consumption among all the catalysts. Fig. 4a illustrates the distribution of the product fractions obtained using different catalysts for the hydrogenation of VR. In the absence of a catalyst, coke formation (9.5 wt%) and gas formation (9.4 wt%) were detected. The Mo SAs–MoS₂ catalyst showed the best coke inhibition performance, with the lowest coke yield of 0.6 wt%, which was lower than that of the Mo SAs (0.8 wt%), and carbon-supported MoS₂ catalysts (C-MoS₂-1 with a coke yield of 6.3 wt% and C-MoS₂-2 with 5.9 wt%), respectively. The liquid product yield and VR conversion (Fig. 4b) could be obtained by analyzing the obtained products in Fig. 4a. Mo SAs–MoS₂ exhibited an excellent liquid product yield and per pass conversion of VR of 92 wt% and 66 wt%, respectively. Mo SAs (Fig. S13 and S14, ESI†) also showed good product yields, but slightly inferior to Mo SAs–MoS₂. C-MoS₂-1 and C-MoS₂-2 with poor activity showed lower liquid product yields.

Fig. 4 Hydrogenation performance and mechanism of Mo SAs–MoS₂ in the VR slurry phase at 698 K. (a) Yield of different products. (b) Liquid product yields, VR conversion, and TOF_T. (c) Anthracene hydrogenation performance tests in the slurry phase, including anthracene conversion and product selectivity. In situ DRIFTS spectra in H₂ for (d) Mo SAs–Mo Cs. (e) Theoretical peaks of the dissociative adsorption of H on the Mo site and S site at 698 K obtained through AIMD simulation. (f) Mo SAs–MoS₂. Comparison of the Mo SAs–Mo Cs precursor and Mo SAs–MoS₂ catalysts. (g) Mass spectrum. (h) H₂-TPR. (i) PPMS.

To further verify the hydrogenation capacity of the prepared catalysts, the total turnover frequency (TOF_T) was calculated and it exhibited a tight relationship with H₂ consumption and the total metal addition. Specifically, Mo SAs–MoS₂ achieved the highest TOF_T of 0.39 s⁻¹, which was higher than that of Mo SAs, C-MoS₂-1 and C-MoS₂-2 of 0.33, 0.02 and 0.022 s⁻¹, respectively. All these hydrogenation performance tests and analyses indicated that the prepared Mo SAs–MoS₂ catalyst have a superior H₂-activation capability (Table S3, ESI†). Compared with Mo SAs and C-MoS₂ catalysts, Mo SAs–MoS₂ showed better hydrogenation activity. Therefore, anthracene was select as a model compound²⁶ for the hydrogenation experiments to further investigate the synergistic effect between Mo SAs and MoS₂. As depicted in Fig. 4c, the anthracene conversions of Mo SAs–MoS₂, Mo SAs, and C-MoS₂-1 catalysts were 96.6, 94.5, and 89.6 wt%, respectively, all higher than the 79.6 wt% achieved without a catalyst. Product analysis exhibited that the Mo SAs–MoS₂ catalyst had higher selectivity for 8-hydroanthracene (AH8) than Mo SAs and C-MoS₂-1, indicating that the synergy between Mo single atoms and MoS₂ promoted a deep hydrogenation of the aromatic hydrocarbons, which could also explain the higher TOFs and different product compositions of Mo SAs–MoS₂. Furthermore, the anthracene hydrogenation with cyclic regeneration of the catalyst was conducted at 425 °C and a 7 MPa H₂ initial pressure (Fig. S15, ESI†). The conversion was reduced from 96.6 wt% to about 92 wt%, while the selectivity for 8-hydroanthracene (AH8) was basically unchanged, which was mainly due to the loss of catalyst during the cycle test. Thus, the Mo SAs–MoS₂ catalyst demonstrated good cycling stability.

Hydrogenation mechanism

To further investigate the synergistic effect between MoS₂ and Mo SAs, in situ characterization was performed to verify the dynamic changes in the active hydrogen species in the reaction. In situ H₂ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was carried out for different temperatures from 50 °C to 450 °C (Fig. 4d and f). S–H and Mo–H stretching vibration peaks were obtained by AIMD simulation at 698 K as theoretical values. The peaks around 739, 1042, 1255, 1478 cm⁻¹ were ascribed to the S–H bonds, and the peaks around 1060, 1255, 1478 cm⁻¹ corresponded to Mo–H bonds.²⁷ The peak b in Mo SAs–MoS₂ was ascribed to S–H species, while peak a attributed to Mo–H species. In addition, peaks c, d, e were ascribed to Mo–H or S–H species. Thermogravimetry (TG) analysis was next performed to analyze the dispersion of H species in the catalysts. The weight loss curves were observed at 357–461 °C and 235–432 °C, respectively (Fig. S16, ESI†), which indicated different products for Mo SAs–Mo Cs and Mo SAs–MoS₂. From the mass spectra (Fig. 4g), more H₂S was detected in Mo SAs–MoS₂ (∼2 × 10⁻¹⁰) than Mo SAs–Mo (close to zero). However, the formation of NH₃ was similar in the two catalysts as 2 × 10⁻⁹ and 1.6 × 10⁻⁹ (Fig. S17, ESI†). Compared with the yields of H₂S and NH₃, the hydrogen species were more concentrated near the S species. As shown in Fig. 4h, H spillover was verified from the H₂-TPR spectra, and the reduction temperature decreased from 456 °C for Mo SAs–Mo Cs to 329 °C for Mo SAs–MoS₂, which indicated that the H₂ was activated to form active H, which adsorbed on to the surface of Mo SAs–MoS₂ during hydrogenation. As a result, the reduction temperature of Mo SAs–MoS₂ was significantly decreased more than that of Mo SAs–Mo Cs, indicating the existence of hydrogen spillover during the hydrogenation.^28–31 The weak adsorption of H₂ by the Mo SAs-Cs precursor indicated that the Mo clusters lacked the ability to improve hydrogen spillover, which made it difficult for hydrogen to migrate and desorb at high temperature. The strong chemisorption (450 °C) of the Mo single-atom site in the Mo SAs–MoS₂ catalyst accelerated the adsorption and dissociation of H₂, while the in situ generation of MoS₂ significantly changed the adsorption behavior of H₂. Furthermore, the surface acceptor site received the H atoms migrated by the Mo single atom to form a weak adsorption species, resulting in the appearance of the low-temperature desorption peak (227 °C).

In addition, the number of vacancies in Mo SAs–MoS₂ was higher than that in Mo SAs–Mo Cs (Fig. S19, ESI†). The induction of active hydrogen was a necessary condition for the formation of oxygen vacancies, which further verified that the H₂ was activated and spilled over to the surface of the catalyst.^32,33 Magnetic measurements are often used to determine the unsaturated electrons in a sample. Here, physical property measurement system (PPMS) tests were performed and the results demonstrated that the unsaturated electron contents of the prepared precursor and catalysts followed the order: Mo SAs–Mo Cs > Mo SAs–MoS₂ > C-MoS₂-1 > Mo SAs (Fig. 4i). Compared with Mo SAs–Mo Cs and Mo SAs, it could be inferred that the unsaturated electrons of the Mo SAs–Mo Cs were more derived from the Mo clusters. Moreover, compared with the Mo SAs–MoS₂ and C-MoS₂-1 catalysts, the single Mo atoms in Mo SAs–MoS₂ were the main origin of the unsaturated electrons, which indicated that the active hydrogen adsorbed on the surface of Mo atoms after activation migrated to the vacancies in MoS₂, further inhibiting the continuous growth of MoS₂. Electrochemical impedance spectroscopy (EIS) was next carried out to explore the conductivity of the catalysts (Fig. 5a). The charge-transfer resistance of the precursor and catalysts were in the order of Mo SAs–Mo Cs < Mo SAs–MoS₂ < C-MoS₂-1 < Mo SAs. A high resistivity corresponds to less unsaturated electrons in the catalyst, which further supports the inference obtained by PPMS.

Fig. 5 Catalytic mechanism investigation. (a) EIS spectra of Mo SAs–Mo Cs, Mo SAs–MoS₂, C-MoS₂-1, and Mo Sas. (b) H₂ adsorption energy and (c) charge density under basic conditions in C-Mo, C-MoS₂, and C-Mo–MoS₂. Yellow and cyan regions represent electron accumulation and depletion. (d) Energy barrier of the hydrogen spillover transition state from the Mo single atom to MoS₂. (e) Anthracene adsorption energy under basic conditions in C-Mo, C-MoS₂, C-Mo–MoS₂. (f) pDOS of C-Mo–MoS₂, C-Mo–MoS₂-AH, and C-Mo–MoS₂-H₂.

Theoretical evidence of hydrogenation for Mo SAs–MoS₂

DFT calculations were performed to investigate the synergistic effect of Mo SAs and MoS₂ for H₂ activation and anthracene adsorption. Several types of Mo species supported on the plane of a C matrix, including C-Mo, C-MoS₂, C-Mo–MoS₂, were selected as reference prototypes to understand the hydrogenation in the slurry phase (Fig. S20–S22, ESI†). The catalytic behaviors of Mo species showed a remarkable correlation with their exposed active sites. As shown in Fig. 5b, the free energy of H₂ adsorption (ΔG_H) for C-Mo–MoS₂ was −5.3 eV, which was significantly higher (in absolute terms) compared to that of C-Mo and C-MoS₂, as −1.5 eV and −1.6 eV, respectively. The much lower ΔG_H indicated that the synergistic effect of Mo SAs and MoS₂ promoted the adsorption and dissociation of H₂. The charge-density-difference plots (Fig. 5c) suggested that substantial charge transfer occurred from the single Mo and the edge of MoS₂ sites to the adsorbed H sites. Significant charge depletion was observed on Mo and S sites, indicating a mixed characteristic of covalent and ionic bonds for the Mo–H and S–H interactions. Hydrogen spillover constrained the marginal growth of MoS₂, and the corresponding transition state in each elementary step was demonstrated in Fig. 5d. The control step of hydrogen spillover exhibited ideally low H-transport barriers of 1.22 eV, which demonstrated a favorable potential for H migration. The MoS₂ in the C basal plane could barely adsorb anthracene. However, anthracene could adsorb on the single Mo site in the C basal plane with a low barrier of −3.2 eV. In comparison, the adsorption of anthracene at the single Mo and MoS₂ co-sites was favorable with clearly lower barriers of −3.5 eV (Fig. 5e). The charge-density-difference plots (Fig. S23, ESI†) suggested that the adsorption process of anthracene on the catalyst surface had a large amount of charge transfer with single atom Mo, while MoS₂ mainly had a steric effect, which increased the adsorption instability of anthracene. Fig. 5f illustrates the calculated density of states (DOS) for the Mo d orbitals. One can see that for the C-Mo–MoS₂ model, the distribution of d states in energy space (indicated by the gray line) was narrow and predominantly situated away from the Fermi level. Conversely, for the C-Mo–MoS₂-H₂ and C-Mo–MoS₂-en systems, the d states (represented by the orange and blue line) were more widespread and made a large contribution near the Fermi level. This implied that the C-Mo–MoS₂ can bond and engage more effectively with H₂ and anthracene and therefore will play a more active role in the hydrogenation of VR.

Conclusions

In summary, we report a robust Mo SAs–MoS₂ coupling catalyst for the slurry-phase hydrogenation of VR. The Mo SAs–MoS₂ catalyst was constructed by a reaction induction transformation strategy. The spontaneous generation of the Mo SAs–MoS₂ coupling catalyst synergistically promotes H₂ activation and enhances the local active hydrogen concentration in the hydrogenation of VR. The yield of liquid product was 92 wt%, and the coke only content was 0.6 wt% with a TOF_T of 0.39 s⁻¹. It was also found that the enhancement of the active H concentration avoided the continuous aggregation and growth and deactivation of the MoS₂ catalyst during the hydrogenation reaction. Theoretical calculations revealed that the synergistic effects of the Mo SAs–MoS₂ coupling catalyst facilitated electron transfer between the Mo species and reactants, reducing the adsorption energy barriers for H₂ and anthracene. The control step for hydrogen spillover exhibited an ideally low H-transport barrier of 1.22 eV, which demonstrates a favorable potential for H migration. This work provides a novel idea for the design of an efficient coupling catalyst for strengthening local active hydrogen concentration in the hydrogenation of VR, and this concept also can be applied to other hydrogenation systems. This study provides a new perspective for designing beneficial coupling hydrogenation catalysts with strengthening the local active hydrogen concentration, which would be beneficial for a series of hydrogenation systems.

Experimental

Chemicals and materials

Vacuum residue (VR) was obtained from Qingdao Refining (Table S1, ESI†). Coal tar pitch (CTP) was obtained from Hebei Hankai Energy Technology Development Co., Ltd. Ammonium molybdate tetrahydrate (AR), melamine (C₃H₆N₆, 99%), magnesium oxide (AR, 98%), ethanol (AR), sulfuric acid (70%), anthracene (GC, 99%), tridecane (GC, 99%), and molybdenum disulfide (99%) were purchased from Aladdin.

Catalyst preparation

Preparation of lamellar magnesium oxide. First, 500 mL deionized water was put into a flask, and 50 g MgO was added. The mixture was stirred evenly, heated and re-fluxed for 8 h (the temperature of the oil bath was set at 100 °C), and then filtered after naturally cooling. The obtained filter cake was wrapped with filter paper and placed in an oven at 110 °C for 12 h. The dried filter cake was ground into powder and calcined in a muffle furnace at 550 °C for 1 h to obtain lamellar MgO.

Preparation of Mo SAs–Mo Cs. First, 1 g CTP and 1 g melamine were dispersed in 40 mL ethanol, and ultrasonicated for 30 min to disperse evenly, and this was marked as solution A. Meanwhile, 600 mg ammonium molybdate tetrahydrate was dissolved in 40 mL deionized water and sonicated for 30 min to completely dissolve, and this was marked as solution B. Then, solutions A and B were mixed under stirring, and 1 g lamellar magnesium oxide was added to the above mixed solution. After stirring for 24 h and drying in the oven at 70 °C, the catalyst precursor was obtained. The catalyst precursor was placed in a crucible and calcined at 650 °C for 4 h in a N₂ atmosphere. Thus, the Mo SAs–Mo Cs precursor was obtained after acid etching the black powder. Here, 0.5 mol L⁻¹ sulfuric acid solution was used for the acid etching process, and this step was maintained for 12 h.

Spontaneous generation of the Mo SAs–MoS₂ catalyst. With the initial H₂ pressure of 7 MPa in the reaction, and rise to 425 °C at the rate of 8 °C min⁻¹. The sulfur-rich environment (sulfur compounds) in the hydrogenation of heavy oil promotes the direct conversion of Mo SAs–Mo Cs into the Mo SAs–MoS₂ catalyst.

Preparation of Mo SAs. First, 1 g CTP and 1 g melamine were dispersed in 40 mL ethanol, and ultrasonicated for 30 min to disperse evenly, and this was marked as solution C. Meanwhile, 30 mg ammonium molybdate tetrahydrate was dissolved in 40 mL deionized water and sonicated for 30 min to completely dissolve, and this was marked as solution D. Then, solutions C and D were mixed under stirring, and 1 g lamellar magnesium oxide was added to the above mixed solution. After stirring for 24 h and drying in the oven at 70 °C, the catalyst precursor was obtained. The catalyst precursor was then placed in a crucible and calcined at 650 °C for 4 h in N₂ atmosphere. Thus, the Mo SAs catalyst was obtained after acid etching the black powder.

Preparation of C-MoS₂-1. First, 1 g CTP and 1 g melamine were dispersed in 40 mL ethanol, and ultrasonicated for 30 min to disperse evenly, and this was marked as solution E. Next, 1 g lamellar magnesium oxide was added to solution E. After stirring for 24 h and drying in the oven at 70 °C, the solid F was obtained. Solid F was then placed in a crucible and calcined at 650 °C for 4 h in a N₂ atmosphere to obtain solid G. Then, solid G was mixed with 200 mg MoS₂ in 100 mL ethyl alcohol, dispersed, and dried, to obtain solid H. Finally, solid H was placed in a crucible and calcined at 650 °C for 4 h in a N₂ atmosphere and the C-MoS₂-1 catalyst was obtained after acid etching the black powder.

Preparation of C-MoS₂-2. First, 1 g CTP and 1 g melamine were dispersed in 40 mL ethanol, and ultrasonicated for 30 min to disperse evenly, and this was marked as solution I. Meanwhile, 200 mg MoS₂ was dissolved in 40 mL deionized water and sonicated for 30 min to completely dissolve, and this was marked as solution J. Then, solution I and J were mixed under stirring, and 1 g lamellar magnesium oxide was added to the above mixed solution. After stirring for 24 h and drying in the oven at 70 °C, the catalyst precursor was obtained. The catalyst precursor was placed in a crucible and calcined at 650 °C for 4 h in a N₂ atmosphere. Thus, the Mo SAs catalyst was obtained after acid etching the black powder.

Catalyst characterization

N₂ physical adsorption/desorption analysis was performed using a Micrometritics ASAP2020 instrument. Inductively coupled plasma-optical emission spectrometry (ICP-OES) measurements were carried out using a Thermo Scientific Icap 6300 system. X-Ray diffraction (XRD) analysis was conducted using a Panalytical X'pert PRO X-ray diffractometer equipped with a Cu Kα radiation source. Scanning electron microscopy (SEM) was performed using a Hitachi S-4800 electron microscope. High-resolution transmission electron microscopy (HRTEM) and high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) with energy-dispersive X-ray spectroscopy (EDS) mapping were performed using a JEOL-2100F electron microscope operating at a working voltage of 200 kV. Aberration-corrected HAADF-STEM was performed on a Titan 80-300 microscope operating at a working voltage of 300 kV. Synchrotron X-ray absorption spectroscopy (XAS) was carried out at the stations of the Shanghai Synchrotron Radiation Facility (SSRF), specifically at Beamlines 11B (BL11B) and 13SSW (BL13SSW). XAS measurements were performed at the Mo K-edge in the fluorescence mode using a Lytle detector. In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) tests were captured using a Bruker VERTEX 70V FTIR spectrometer. X-Ray photoelectron spectroscopy (XPS) was performed using an ESCALAB MK II spectrometer equipped with an Al Kα (1486.6 eV) monochromated photon source and two ultrahigh vacuum (UHV) chambers. Electron spin resonance (ESR) spectroscopy was carried out using a Bruker EMX Plus spectrometer. Physical property measurement system (PPMS) tests were performed using an America-Quantum Design-PPMS 16T instrument.

Catalyst evaluation

The catalytic performance of the catalysts in hydrocracking was evaluated using a batch mode experiment in a 500 mL autoclave. VR (100 g) was used as the feedstock. The experiments were carried out at a reaction temperature of 425 °C and an initial hydrogen pressure of 7 MPa, utilizing Mo SAs–MoS₂ catalyst at a concentration of 200 ppm, calculated based on the weight of molybdenum. The reaction was maintained under a stirring speed of 500 rpm for a duration of 1 h. After the reaction, the autoclave was brought down to room temperature, following which the gas products were expelled. The toluene-insoluble components within the product were identified as coke deposits and the catalyst. The fractionation distribution of the liquid product was analyzed via the gas chromatography-simulated distillation (GC-SD) methodology. This GC-SD methodology segregates the liquid product into four distinct fractions: unconverted oil (UCO, boiling point >500 °C), vacuum gas oil (VGO, boiling range 350–500 °C), diesel fraction (boiling range 180–350 °C), and naphtha fraction (boiling point <180 °C). The mass of the vapor phase output was ascertained by deducting the weight of the liquid product and the toluene-insoluble components from the total reactor volume. The anthracene hydrogenation activity of the Mo SAs–MoS₂ catalysts was evaluated in a batch-wise manner using a 200 mL autoclave setup. In the reaction, Mo SAs–MoS₂ (0.5 g), anthracene (4 g) and tridecane (50 g) served as the reactants, with the catalyst being simultaneously incorporated into the reaction mixture. The experiments aimed at anthracene hydrogenation were performed under a reaction temperature of 425 °C and an initial hydrogen pressure of 7 MPa. The reaction conditions were maintained with a stirring speed of 700 rpm for 4 h. After the reaction, the product was cooled to room temperature using the autoclave. The products were analyzed using liquid chromatography and included octahydroanthracene (AH8), tetrahydroanthracene (AH4), and dihydroanthracene (AH2).

The calculations of the liquid product yield and VR conversion were performed using the following equations:

Yield of liquid products (wt%) = Yields of (naphtha + diesel + VGO + UGO)

TOF_T was calculated using the following equation:

The conversion and selectivity for the hydrogenation products of anthracene were calculated using the following equations:

Computational details

All the computations using density functional theory (DFT) were conducted with the Vienna Ab initio Simulation Package (VASP) version 5.4.4.^34,35 The Perdew–Burke–Ernzerhof (PBE) functional, based on the generalized gradient approximation (GGA), was employed to describe the exchange–correlation energy.³⁶ The projector-augmented wave (PAW) pseudopotential was utilized to account for the interaction between ions and electrons. The Grimme approach (DFT-D3) was applied to correct for weak van der Waals interactions. The kinetic energy cut-off for the plane-wave expansion was set to 450 eV. A 3 × 3 × 1 Monkhorst–Pack k-point grid was used for efficient structure optimizations. The convergence criteria for the energy and forces were set to 1 × 10⁻⁵ eV and 0.03 eV Å⁻¹, respectively. The identification of the transition states was achieved using the climbing image nudged elastic band method (CI-NEB). A meticulously constructed model of single-layer graphene, exhibiting periodic repetition, was established with a unit cell dimension of 6 × 6 and a vacuum slab height of 25 Å.

The adsorption energies of hydrogen were calculated using the formula ΔG = G_slab+2H − G_slab − G_H2(g), where G_slab+2H, G_slab, and G_H2(g) represent the Gibbs free energy of the catalyst model slab with adsorbed hydrogen, the slab, and gaseous hydrogen, respectively. The Gibbs free energy was calculated according to the equation: G = E + ZPE − TS where E, ZPE, and S are the DFT-calculated energies, zero-point energies, and entropy, respectively.

Author contributions

Conceptualization: G. S., Y. P. Methodology: G. S., P. X., Y. L., H. S. Investigation: X. Z., F. T., Z. G. Visualization: C. Y., J. L., B. L. Supervision: Y. L. Z. L. Writing – original draft: G. S., Z. L. Writing review & editing: Y. P.

Conflicts of interest

All the authors declare they have no competing interests.

Data availability

The data supporting this article have been included as part of the ESI.†

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (22478432 and 22108306), Shandong Provincial Natural Science Foundation (ZR2024JQ004).

Notes and references

1. E. T. C. Vogt and B. M. Weckhuysen, Chem. Soc. Rev., 2015, 44, 7342–7370.
2. R. Prajapati, K. Kohli and S. K. Maity, Fuel, 2021, 288, 119686.
3. Y. Ma, W. Wu, J. Zhang, H. Zhang, Z. Cai, Y. Cao, K. Huang and L. Jiang, Ind. Eng. Chem. Res., 2023, 62, 15459–15468.
4. R. Qu, K. Junge and M. Beller, Chem. Rev., 2023, 123, 1103–1165.
5. C. Wu, X. Chen, L. Tang, Q. Wei, J. Liang and L. Wang, ACS Appl. Mater. Interfaces, 2021, 13, 10855–10869.
6. K. H. Kang, N. T. Nguyen, D. V. Pham, P. W. Seo, N. Kang, C. W. Lee, M.-C. Chung, C.-H. Kwak, I. Ro, Y.-P. Jeon and S. Park, J. Catal., 2021, 402, 194–207.
7. K. D. Kim and Y. K. Lee, J. Catal., 2019, 369, 111–121.
8. S. H. Kim, K. D. Kim and Y. K. Lee, J. Catal., 2017, 347, 127–137.
9. S. H. Kim, K. D. Kim, D. Lee and Y. K. Lee, J. Catal., 2018, 364, 131–140.
10. Z. Guan, M. Wang, G. Sun, X. Xue, Y. Chen, Y. Zhang, T. Zhao, H. Shi, C. Feng, Y. Pan and Y. Liu, Fuel, 2024, 357, 129887.
11. A. V. Bekker and I. Livk, Ind. Eng. Chem. Res., 2011, 50, 3464–3474.
12. G. Sun, D. Liu, M. Li, S. Tao, Z. Guan, Y. Chen, S. Liu, Q. Du, H. Guo, X. Yuan, X. Zhang, H. Zhu, B. Liu and Y. Pan, Sci. Bull., 2023, 68, 503–515.
13. G. Sun, D. Liu, H. Shi, J. Li, L. Yang, F. Tian, Y. Cui, C. Wang, F. Li, T. Zhao, H. Zhu, B. Liu, Y. Chai, Y. Liu and Y. Pan, ACS Catal., 2024, 14, 3208–3217.
14. L. Zhang, M. Zhou, A. Wang and T. Zhang, Chem. Rev., 2020, 120, 683–733.
15. Y. Zhang, Y. Cheng, X. Wang, Q. Sun, X. He and H. Ji, ACS Catal., 2022, 12, 15091–15096.
16. H. Jin, S. Sultan, M. Ha, J. N. Tiwari, M. G. Kim and K. S. Kim, Adv. Funct. Mater., 2020, 30, 2000531.
17. Q. Feng, S. Zhao, Q. Xu, W. Chen, S. Tian, Y. Wang, W. Yan, J. Luo, D. Wang and Y. Li, Adv. Mater., 2019, 31, 1901024.
18. R. Huang, M. Xia, Y. Zhang, C. Guan, Y. Wei, Z. Jiang, M. Li, B. Zhao, X. Hou, Y. Wei, Q. Chen, J. Hu, X. Cui, L. Yu and D. Deng, Nat. Catal., 2023, 6, 1005–1015.
19. M. Li, G. Sun, Z. Wang, X. Zhang, J. Peng, F. Jiang, J. Li, S. Tao, Y. Liu and Y. Pan, Adv. Mater., 2024, 36, 2313661.
20. R. Chen, S. Chen, L. Wang and D. Wang, Adv. Mater., 2024, 36, 2304713.
21. C. Yue, C. Feng, G. Sun, N. Liu, H. Hao, W. Bao, X. Zhang, F. Sun, C. Zhang, J. Bi, Y. Zhou, H. C. Chen, Y. Pan, D. Sun and Y. Lu, Energy Environ. Sci., 2024, 17, 5227–5240.
22. Y. Wang, M. Qi, M. Conte, Z. Tang and Y. Xu, Angew. Chem., Int. Ed., 2024, e202407791.
23. X. Xu, X. Li, W. Lu, X. Sun, H. Huang, X. Cui, L. Li, X. Zou, W. Zheng and X. Zhao, Angew. Chem., Int. Ed., 2024, 63, e202400765.
24. W. Ni, J. L. Meibom, N. U. Hassan, M. Chang, Y. C. Chu, A. Krammer, S. Sun, Y. Zheng, L. Bai, W. Ma, S. Lee, S. Jin, J. S. Luterbacher, A. Schüler, H. Chen, W. E. Mustain and X. Hu, Nat. Catal., 2023, 6, 773–783.
25. Z. Li and C. Yu, Nano. Mater., 2024, 185–224.
26. P. Zhang, M. Guo, F. Wang, Q. Tang, H. Ge, S. Guo, L. Zhou, X. Li, J. Dong, M. Tang and L. Liu, J. Mater. Sci., 2024, 59, 21027–21039.
27. R. M. Ramadan, M. S. A. Hamaza and S. A. Ali, J. Cood. Chem., 1998, 43, 31–39.
28. H. Kang, L. Zhu, S. Li, S. Yu, Y. Niu, B. Zhang, W. Chu, X. Liu, S. Perathoner, G. Centi and Y. Liu, Nat. Catal., 2023, 6, 1062–1072.
29. M. Xiong, Z. Gao and Y. Qin, ACS Catal., 2021, 11, 3159–3172.
30. K. Liu, P. Yan, H. Jiang, Z. Xia, Z. Xu, S. Bai and Z. C. Zhang, J. Catal., 2019, 369, 396–404.
31. Q. Yu, T. Yu, H. Chen, G. Fang, X. Pan and X. Bao, J. Energy Chem., 2020, 41, 93–99.
32. Y. Yang, S. Zhao, F. Bi, J. Chen, Y. Li, L. Cui, J. Xu and X. Zhang, Cell Rep. Phys. Sci., 2022, 3, 101011.
33. L. Liccardo, M. Bordin, P. M. Sheverdyaeva, M. Belli, P. Moras, A. Vomiero and E. Moretti, Adv. Funct. Mater., 2023, 33, 2212486.
34. G. Kresse and J. Furthmüller, Comput. Mater. Sci., 1996, 6, 15–50.
35. G. Kresse and J. Furthmüller, Phys. Rev. B: Condens. Matter., 1996, 54, 11169.
36. J. P. Perdew, K. Burke and Y. Wang, Phys. Rev. B: Condens. Matter Mater. Phys., 1998, 54, 16–533.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ey00086f

---