Role of sulfur vacancies in MoS₂ monolayers in stabilizing Co atoms for efficient CO oxidation

Abstract

By performing first-principles calculations, a MoS₂ monolayer with a Co atom doped at the sulfur defect (Co-^SMoS₂) was investigated as a single-atom catalyst (SAC) for CO oxidation. The Co atom is strongly constrained at the S-vacancy site of MoS₂ without forming clusters by showing a high diffusion energy barrier, ensuring good stability to catalyze CO oxidation. The CO and O₂ adsorption behavior on Co-^SMoS₂ surface and four reaction pathways, namely, the Eley–Rideal (ER), Langmuir–Hinshelwood (LH), trimolecular Eley–Rideal (TER) as well as the New Eley–Rideal (NER) mechanisms are studied to understand the catalytic activity of Co-^SMoS₂ for CO oxidation. The CO oxidation is more likely to proceed through the LH mechanism, and the energy barrier for the rate-limiting step is only 0.19 eV, smaller than that of noble metal-based SACs. Additionally, the NER mechanism is also favorable with a low energy barrier of 0.26 eV, indicating that the Co-^SMoS₂ catalyst can effectively promote CO oxidation at low temperatures. Our investigation demonstrates that the S-vacancy of MoS₂ plays an important role in enhancing the stability and catalytic activity of Co atoms and Co-^SMoS₂ is predicted to be a promising catalyst for CO oxidation.

---

1. Introduction

Carbon monoxide (CO), originating from deficient combustion of carbon-containing fuels (such as coal and oil) from automobiles and industrial processes, is a widely distributed toxic gas in the atmosphere and can severely poison noble metal catalysts in fuel cells.^1–3 The oxidation of CO to CO₂ at low temperature is an effective strategy to relieve the problems caused by increased CO emission.^4–7 Although the oxidation of CO is exothermic, the slow kinetics of this reaction demands the development of efficient catalysts to reduce the activation barrier.⁸ In the past decades, noble metal catalysts, such as Pt,^9,10 Au,^11,12 Pd,^13,14 and Rh^15,16 have been extensively explored and exhibit good catalytic activity for CO oxidation. However, these catalysts are expensive and typically require high reaction temperatures to operate efficiently, which prevents their large-scale commercial applications.¹³ Therefore, low-cost and efficient non-noble metal-based catalysts are highly demanded.

In 2011, the preparation of homogeneous single Pt atoms on FeO_X was first reported, and the Pt/FeO_X catalysts showed good catalytic performance for CO oxidation.¹⁷ Thereafter, single-atom catalysts (SACs) have been rapidly developed as an important class of catalysts by showing good catalytic activity and selectivity for many reactions including CO oxidation, ^18,19 oxygen reduction reactions (ORR),^20,21 oxygen evolution reactions (OER),^22,23 and hydrogen evolution reactions (HER).^24,25 In contrast to conventional bulk metal catalysts, SACs are supported catalysts containing individual atoms isolated from each other as active centers, maximizing the utilization efficiency of metal atoms and reducing the cost.²⁶ However, the isolated single metal atoms prone to be aggregated because of their high surface free energy and stable support materials are crucial to obtaining stable SACs with finely dispersed metal centers.

Recently, 2D materials have been widely studied as SAC supports because of their large surface area, tunable electronic properties and good stability, and some 2D SACs have been demonstrated to be promising catalysts for CO oxidation.^27–34 Among them, 2D transition metal dichalcogenides (TMDCs) represented by MoS₂ monolayers have drawn broad research interests because of their good stability and high experimental accessibility.^35–38 Previous studies have demonstrated that doping transition metals on MoS₂ monolayer can effectively tune the electronic structure and magnetic properties of MoS₂,³⁹ significantly enhancing the interaction between the catalyst and the adsorbed CO and O₂ to facilitate CO oxidation.^40–42 For instance, MoS₂ monolayers with Co-doped on the pristine surface was demonstrated to show high catalytic activity for CO oxidation.⁴³ However, aggregation of Co on pristine MoS₂ will be inevitable due to the weak interaction between Co and the defect-free surface. In fact, the embedding of non-precious metals in the S vacancy of MoS₂ can be experimentally feasible,^44,45 because when 2D MoS₂ was prepared using mechanical exfoliation or chemical vapor deposition techniques, S vacancies were the predominant defects and can even be controlled by using low-energy argon sputtering or electron irradiation techniques.^46–48 Recently, Liu et al. synthesized MoS₂ monolayer with cobalt-embedded at the sulfur vacancies by mixing chemically exfoliated MoS₂ monolayer containing S vacancies with thiourea-like cobalt complexes,⁴⁹ which exhibited high activity and stability for hydrodeoxygenation reaction and formaldehyde oxidation.⁵⁰ Inspired by the good stability and high synthesizing feasibility of MoS₂ monolayers with Co embedded in the sulfur vacancies, it is rather appealing to investigate their catalytic performance for CO oxidation, which has not been explored before.

In this work, we performed first-principles calculations to study the potential of MoS₂ monolayers with individual Co atom embedded in the S vacancy (Co-^SMoS₂) as SACs for CO oxidation. The stability of Co-^SMoS₂ was first verified by estimating the diffusion of Co from the S vacancy to different adjacent surface sites. Then, the adsorption of CO and O₂ on Co-^SMoS₂ was examined in comparison with that on defect-free MoS₂ monolayer, defect-free MoS₂ monolayer with a single Co atom decorated on the surface (Co/MoS₂) and MoS₂ monolayer containing S vacancies (MoS₂-Sv). Four CO oxidation mechanisms on the Co-^SMoS₂, including, Eley–Rideal (ER), New Eley–Rideal (NER), Langmuir–Hinshelwood (LH) and trimolecular Eley–Rideal (TER) mechanisms, were systematically explored. The results indicate that Co-^SMoS₂ can facilitate CO oxidation through the LH and NER mechanisms by surmounting small energy barriers and show superior stability and catalytic activity than other examined MoS₂ based structures because of the strong interaction between Co and sulfur vacancy, thus is predicted to be a promising catalyst for CO oxidation.

2. Computational methods

All calculations were carried out with the projected augmented wave (PAW) potential of spin-polarized density functional theory (DFT), as implemented in the Vienna Ab initio Simulation Package (VASP).^51–53 The generalized gradient approximation (GGA) method with Perdew–Burke–Ernzerhof (PBE) was adopted to depict exchange–correlation potential.⁵⁴ The van der Waals (vdWs) interactions between the adsorbing molecules and the monolayer were depicted by using the DFT-D3 method proposed by Grimme.⁵⁵ The plane-wave cutoff energy was set as 420 eV and the Monkhorst–Pack grid of 3 × 3 × 1 and 5 × 5 × 1 k-points grid were used for the geometry optimization and the calculation of electronic properties. The system is considered to be self-consistent when the total energy and the magnitude of the force on each atom are below 10⁻⁵ eV and 0.01 eV Å⁻¹, respectively. The vacuum layer of the MoS₂ monolayer was set to be 20 Å to avoid interlayer interactions. The minimum energy paths for CO oxidation reaction were searched by the climbing image nudged elastic band (CI-NEB) and dimer method,^56–58 and the transition state (TS) geometries were obtained after vibrational analysis. The lattice parameter of MoS₂ monolayer was optimized to be 3.17 Å, which agrees well with the results of previously experimental and theoretical studies.^43,59 The binding energy (E_b) of Co atom in Co-^SMoS₂ was calculated by

E_b = E_Co-^SMoS2 − E_MoS2-Sv − E_Co (1)

where E_Co-^SMoS2, E_MoS2-Sv, and E_Co are the total energy of the defective MoS₂ embedded with a Co atom, MoS₂ with a S vacancy and the free Co atom, respectively.

The adsorption energy (E_ads) of different adsorbates on different substrates was calculated according to

E_ads = E_sub+gas − E_sub − E_gas (2)

where E_sub+gas, E_sub and E_gas represent the total energies of the substrate adsorbed with gas molecule, the bare substrate and the free gas molecule, respectively. A more negative E_ads value indicates a more favorable adsorption of the gas molecule.

To investigate whether the CO oxidation performance is affected by the doping concentration of Co, we constructed a 2 × 2, 3 × 3, 4 × 4 and 5 × 5 supercell containing one Co atom embedded at the S vacancy, corresponding to the atomic embedding concentration of 25%, 11%, 6% and 4%, respectively. However, it turned out that these four supercells gave identical E_b values of Co and E_ads for CO and O₂ (Table S1†), indicating that the concentration of Co shows negligible influence on their catalytic activity. Therefore, considering the computational accuracy and efficiency, the 4 × 4 MoS₂ supercell is used for all catalysis calculations unless stated otherwise.

The reaction energy barriers (E_bar) and reaction energy (ΔE) for the elementary steps of CO oxidation reaction was defined as

E_bar = E_TS − E_IS (3)

ΔE = E_FS − E_IS (4)

where E_TS, E_IS and E_FS denote the energy of the TS, initial state (IS) and final state (FS), respectively. A more positive ΔE value indicates a more endothermic reaction process, while a negative value represents an exothermic reaction.

3. Results and discussion

3.1 Structure and stability of Co-^SMoS₂

The structural and electronic properties of Co-^SMoS₂ were first studied. As shown in Fig. 1a, the Co atom is embedded at the S vacancy forming three covalent bonds with its adjacent Mo atoms and the average bond length of Co–Mo is 2.51 Å. The E_b of Co atom is −4.39 eV, indicating a strong interaction of Co atom with the neighboring Mo atoms, which is stronger than that on the defect-free MoS₂ surface (−3.28 eV). In view of the cohesion energy of bulk Co (−4.39 eV),⁶⁰ it can be expected that the embedded Co atoms are less likely to aggregate to form large clusters on defective MoS₂ compared to the defect-free surface. To gain a deep understanding of the stability of Co-^SMoS₂, we estimated the migration of Co atom from the S vacancy to a nearby hollow site. As shown in Fig. 1b, the calculated diffusion process is endothermic (1.26 eV) with a high diffusion energy barrier of 1.71 eV. In contrast, the E_b of Co atom on two different surface sites of defect-free MoS₂ only differs in 0.25 eV and the diffusion energy barrier of Co is 1.08 eV. The results suggest that the S vacancies on MoS₂ play an important role in stabilizing the Co atoms and inhibiting their further aggregation.

Fig. 1 (a) Top and side views of optimized Co-^SMoS₂. (b) The IS, TS and FS geometries of Co-^SMoS₂ (blue line) and Co/MoS₂ (black line) along the diffusion path for Co atom migrating from the S vacancy and Mo-top site to the nearby hollow site, respectively. (c) The PDOS of Co-3d (red curve) and Mo-4d (blue curve) states for Co-^SMoS₂. The Fermi level is set to zero. (d) CDD for Co-^SMoS₂, yellow and cyan areas indicate charge accumulation and depletion, respectively. The isosurface value is set as 0.0018 e/bohr³.

The spin polarized partial density of states (PDOS) projected on the Co-3d and adjacent Mo-4d orbitals were plotted in comparison with the total DOS to understand the electronic structure of Co-^SMoS₂. As shown in Fig. 1c, the embedding of Co atom induces impurity states around the Fermi energy level, contributing to a reduced band gap (0.20 eV) of Co-^SMoS₂ compared to the pristine MoS₂ monolayer. The pronounced hybridization between the Co-3d orbital and the adjacent Mo-4d orbitals around the Fermi energy level is responsible for the strong interaction between Co and defective MoS₂. Bader charge analysis⁶¹ shows that there is about 0.14|e| charge transferred from the Co atom to the adjacent Mo atoms, which can be visualized by the charge density difference (CDD) analysis. As illustrated in Fig. 1d, there is obvious electron depletion (cyan) and electron accumulation (yellow) at the Co and the neighboring Mo atoms, respectively. The strong interaction and significant charge reallocation between Co and defective MoS₂ will favor the good stability and high activity of Co-^SMoS₂ for efficient CO oxidation.

3.2 Activation of gas molecules on different MoS₂ monolayers

Before exploring the mechanism of CO oxidation on Co-^SMoS₂ surfaces, the adsorption and co-adsorption of the reactants (CO and O₂) were first examined. The most preferred adsorption site for each molecule was decided by considering different adsorption configurations. The CO molecule prefers to be adsorbed at the Co site of Co-^SMoS₂ with the end-on configuration, as exhibited in Fig. 2a. The calculated E_ads is −1.78 eV and the bond length of Co–C is 1.79 Å, indicating a strong interaction between the adsorbed CO and the Co-^SMoS₂. Bader charge analysis indicates about 0.20|e| charge transferred from Co-^SMoS₂ to CO and CDD analysis (Fig. 2b) illustrates the electron accumulation at the Co–C bond. The PDOS analysis (Fig. S2a, ESI†) shows that there is obvious hybridization between the Co-3d orbital and the C-2p and O-2p orbitals near the Fermi level. The significant charge transfer and strong hybridization between Co atoms and the adsorbed CO molecule indicate that the CO molecule can been effectively activated by the Co, leading to elongated C–O bond from 1.14 to 1.16 Å.

Fig. 2 Top and side views of the optimized structure and CDD of CO (a and b), O₂ (c and d) CO + O₂ (e and f) and 2CO (g and h) adsorbed Co-^SMoS₂. Yellow and cyan areas denote charge accumulation and depletion, respectively. The isosurface value is set as 0.002 e/bohr³.

The most stable adsorption configuration of O₂ on Co-^SMoS₂ can be seen in Fig. 2c. It can be found that the O₂ molecule is adsorbed at the Co site in a side-on manner, forming two Co–O bonds with the bond length of 1.89 Å, and the calculated E_ads of −1.64 eV is a little lower than that of CO. The Bader charge analysis shows that there is about 0.46|e| charge transferred from the Co-^SMoS₂ to O₂, which will occupy the 2π* orbital of O₂ and contribute to the elongation of the O–O bond from 1.23 to 1.36 Å. As shown in Fig. 2d, there is significant electron accumulation at the adsorbed O₂. These results indicate that the O₂ can be highly activated at the Co site of Co-^SMoS₂. Moreover, the hybridization (Fig. S2b, ESI†) between the Co-3d state and the O₂-2p state also suggests the significant interaction of Co atom with the adsorbed O₂ molecule.

The adsorption of CO and O₂ on defect-free MoS₂, Co/MoS₂ and MoS₂-Sv surfaces (Fig. S3 and Table S2, ESI†) were also examined as comparisons. It can be found that the adsorption of CO and O₂ on defect-free MoS₂ is rather weak and the E_ads is as small as −0.15 and −0.09 eV, respectively. Thus, CO and O₂ can only be physisorbed on defect-free MoS₂ monolayer. On the MoS₂-Sv surface, CO and O₂ molecules are adsorbed at the S vacancy in an end-on and side-on manner, respectively. The C–O bond of CO and the O–O bond of O₂ is elongated to 1.16 and 1.44 Å and the E_ads is −1.24 and −2.14 eV, respectively. Therefore, the S vacancy of MoS₂-Sv prefers to be occupied by O₂ instead of CO. On the Co/MoS₂ surface, the E_ads of CO and O₂ is −2.30 and −1.90 eV, respectively, and the C–O and O–O bond length increase to 1.16 and 1.35 Å, respectively. It can be found that surface defects play important roles in enhancing the ability of MoS₂ monolayer to activate surface molecules and different kinds of defects contribute to different activation of CO and O₂, which will result in different catalytic performances for CO oxidation.

To better reveal the mechanism of CO oxidation reaction on the surface of different catalysts, the co-adsorption of the reactants is also investigated because the preferred adsorption manner can determine the pathway of CO oxidation. Fig. 2e and g show the most stable co-adsorption configurations of CO + O₂ and two CO molecules on Co-^SMoS₂, respectively. The co-adsorption energy of CO + O₂ is calculated to be −2.04 eV, more negative than that of the isolate adsorption of CO or O₂, indicating that co-adsorption of CO + O₂ molecules on Co-^SMoS₂ is more favored. The co-adsorption of two CO molecules is also examined which turns out to be more favorable than that of one CO molecule alone, where a V-type OCCoCO structure was formed with an E_ads of −2.81 eV. There is an obvious charge accumulation and depletion between the Co atom and the adsorbed CO + O₂ or 2CO molecules (Fig. 2f and h). Thus, in terms of E_ads, the co-adsorption of 2CO molecules is even more feasible than that of CO + O₂ molecules on Co-^SMoS₂.

For the defect-free MoS₂, the adsorption of CO + O₂ and 2CO remains weak with the E_ads of −0.28 eV and −0.33 eV, respectively (Fig. S4a and Table S2, ESI†); for the MoS₂-Sv surface, the second gas molecule is difficult to be adsorbed after the adsorption of one molecule of CO or O₂ on the S vacancy (Fig. S4b, ESI†), indicating that the S vacancy tends to be oxidized at the exposure of CO or O₂, which will cause catalyst deactivation. Thus we further examined the robustness of MoS₂-Sv by estimating the feasibility of removing the O* species at the S vacancy using a molecule of CO (Fig. S5, ESI†). The results show that this step (CO + O* → CO₂) requires to conquer a high energy barrier of 2.17 eV and this elementary reaction is endothermic with the reaction energy of 0.83 eV, indicating that the O* species is difficult to be removed and the S vacancy will be definitely poisoned. The co-adsorption configurations of CO + O₂ and 2CO on Co/MoS₂ are similar to those on Co-^SMoS₂ and exhibit the E_ads of −2.67 and −3.29 eV, respectively (Fig. S4c, ESI†). Since defect-free MoS₂ shows weak adsorption of the reactants and MoS₂-Sv tends to be oxidized by the reactants, they are both unfavorable for subsequent CO oxidation, and thus are excluded from potential catalysts for further discussion. We only examined the detailed CO oxidation process on Co-^SMoS₂ and Co/MoS₂ in subsequent calculations unless stated otherwise.

The E_ads of CO₂ on Co-^SMoS₂ and Co/MoS₂ was then examined since CO₂ is the final product of CO oxidation, which turns out to be −0.48 eV for both cases (Fig. S3c, d and Table S2, ESI†). This indicates that the adsorption of CO₂ molecule on Co-^SMoS₂ and Co/MoS₂ is feeble and can be desorbed from the surface site at room temperature. Furthermore, when the free O₂ or CO molecules approach to the Co-^SMoS₂ with CO₂ pre-adsorbed, the CO₂ is automatically desorbed from the catalyst (Fig. S6, ESI†). Thus, even in the presence of abundant CO₂, O₂ and CO molecules can still reoccupy the active site to promote the next step of CO oxidation.

3.3 CO oxidation over Co-^SMoS₂

Generally, CO oxidation reactions can occur through three traditional pathways, namely, the ER, LH, and TER mechanisms,^28,33 which are determined by the initial adsorption configuration of the reactant molecules. For the ER mechanism, O₂ is pre-adsorbed at the reaction site, following by the approaching of CO to form the CO₃* intermediates (IM) or directly to produce a CO₂, leaving a O* species. The LH mechanism is facilitated by the co-adsorption of CO and O₂ molecules at the reaction site to form peroxide-like OOCO* complexes, which then dissociate into CO₂ and O*. The TER mechanism will proceed when the O₂ reacts with the co-adsorbed bi-molecular CO to form OCOOCO* (IM), which subsequently turns into two CO₂ molecules. As mentioned in Section 3.2, the adsorption energy of CO (−1.78 eV) and O₂ (−1.64 eV) on Co-^SMoS₂ are similar, so that different reactants may be adsorbed at the active site when the reactant concentrations vary, and we next discuss the possible reaction mechanisms for different reactant concentrations.

3.3.1 ER mechanism. If the concentration of O₂ is larger than that of CO, CO oxidation will proceed via the ER mechanism. In the first step, the energetically preferred O₂ adsorption structure was chosen as the IS (Fig. 3). As the CO approaches to the Co-^SMoS₂ with O₂ preadsorbed (TS1), the O1–O2 bond of O₂ will be activated and then broken after CO is combined, forming a carbonate-like CO₃* (IM). This step (CO + O₂* → CO₃*) on Co-^SMoS₂ is exothermic with the reaction energy of −3.49 eV and the energy barrier is as small as 0.18 eV. The C–O bond of CO in IM is extended to 1.21 Å and the C–O1/O2 bond length is about 1.35 Å. Meanwhile, the Co–O1 and Co–O2 bond distance is decreased to 1.92 and 1.86 Å, respectively. In the second step, with elongation of the C–O1 and Co–O1 bonds, the IM decomposes into CO₂ and O* (FS1) through TS2. This step (CO₃* → CO₂ + O*) is endothermic by 0.33 eV and shows an energy barrier of 1.03 eV. Then the first CO₂ molecule is formed and released, leaving an O* species stabilized on the Co site (FS1). Subsequently, the second CO molecule approaches to the O* species to form a second CO₂ (CO + O* → CO₂) through TS3 with the reaction energy of −2.19 eV and a very small energy barrier of 0.06 eV. It can be found that the second step (CO₃* → CO₂ + O*) is the rate-determining step (RDS) and IM is more stable than IS and FS1, which is thermodynamically unfavorable, and the high energy barrier makes it kinetically infeasible to proceed through the ER mechanism.

Fig. 3 Reaction pathway for CO oxidation on Co-^SMoS₂ through the ER mechanism.

3.3.2 NER mechanism. Recently, a new mechanism based on the pre-adsorption of O₂, namely the NER mechanism has been proposed for CO oxidation,^62,63 which indicates that two CO molecules can simultaneously attack the adsorbed O₂, facilitating the breakage of the O–O bond to produce two CO₂ molecules. Here, the NER mechanism was examined by using the configuration of two physisorbed CO molecules on the pre-adsorbed O₂ over Co-^SMoS₂ as the IS (Fig. 4). The initial distance between the C atom of CO and O of O₂ is about 3.17 Å and the O–O bond length is 1.36 Å. As the distance between CO and O₂ reduces gradually, the O–O bond is elongated from 1.36 to 2.49 Å, leading to the bond breakage and the formation of two new C–O bonds, forming a pentagonal ring of OOCCOO* (IM) through TS1. This step (2CO + O₂* → OOCCOO*) is exothermic with the reaction energy of −4.05 eV and shows a rather low energy barrier of 0.08 eV. As the C–C bond breaks, the IM is segregated into two CO₂ molecules (FS) by overcoming an energy barrier of 0.26 eV (TS2) and the reaction energy of this step (OOCCOO* → 2CO₂) is −1.30 eV, which turns out to be the RDS along the NER mechanism. After two CO₂ molecules are released, the Co-^SMoS₂ catalyst can be refreshed for further CO oxidation cycles.

Fig. 4 Reaction pathway for CO oxidation on Co-^SMoS₂ through the NER mechanism.

3.3.3 LH mechanism. CO oxidation will proceed through the LH mechanism if the concentrations of CO and O₂ are about equal. As discussed above, the co-adsorption energy of CO + O₂ molecules on Co-^SMoS₂ (−2.04 eV) is more negative than that of the isolated adsorption of CO or O₂, indicating the high feasibility of the LH mechanism. The geometric structure and energy profile of different intermediates of the LH mechanism are presented in Fig. 5. The reaction is initiated with the most favorable co-adsorption of O₂ and CO on Co-^SMoS₂ (IS1), where the O₂ and CO are both adsorbed in an end-on manner. The bond length of O1–O2 in O₂ and C–O in CO is slightly elongated to 1.29 and 1.16 Å, respectively, indicating that both CO and O₂ molecules are activated. Then, CO and O₂ are rotated with O₁ gradually approaching the C atom, contributing to the shortened distance between the CO and O₂ molecules (TS1). After overcoming a small energy barrier of 0.19 eV, a carbonate-like OOCO* species (IM) is formed and the O1–O2 bond is extended to 1.53 Å while the C–O1 bond length is shortened from 2.94 to 1.34 Å. This step (CO* + O₂* → OOCO*) is exothermic with a reaction energy of −0.02 eV. The O₁–O₂ bond is continuously elongated, forming the first CO₂ molecule through TS2, leaving a O* species attached to the Co atom (FS1). This step (OOCO* → CO₂ + O*) is exothermic by showing a reaction energy of −2.79 eV and the energy barrier is 0.07 eV. Then, the step of generating the second CO₂ (CO + O* → CO₂) is consistent with that of the ER mechanism with the reaction energy of −2.19 eV and an energy barrier of 0.06 eV. The RDS is the step of IM generation with a rather small energy barrier of 0.19 eV for the LH mechanism, indicating that CO oxidation can proceed favorably through this pathway.

Fig. 5 Reaction pathway for CO oxidation on Co-^SMoS₂ via the LH mechanism.

3.3.4 TER mechanism. When the concentration of CO is larger than that of O₂, CO oxidation will proceed through the TER mechanism. Since the co-adsorption energy of two CO molecules (−2.81 eV) are also more negative than that of the separated adsorption of CO and O₂, the TER mechanism can be also feasible. Fig. 6 shows the reaction diagram of CO oxidation via the TER mechanism. We took the configuration of O₂ molecules above the two co-adsorbed CO molecules as the IS1, where the bond lengths between two O of O₂ and two C of two CO are 4.02 Å and 3.70 Å, respectively. As the O₂ molecule approaches to the co-adsorbed CO molecules, the O–O bond is elongated from 1.23 to 1.49 Å, forming a pentagonal ring structure of OCOOCO* (IM) through TS1. This step (O₂ + 2CO* → OCOOCO*) is exothermic by showing a reaction energy of −0.26 eV and an energy barrier of 0.75 eV. Then, as the O–O bond breaks, IM dissociates to form two CO₂ molecules (OCOOCO*→ 2CO₂) through TS2 by overcoming an energy barrier of 0.51 eV, and the reaction is exothermic with a reaction energy of −4.21 eV. It is noticed that the IM generation step is the RDS of the TER mechanism, which shows a higher energy barrier (0.75 eV) than the LH (0.19 eV) and NER mechanisms (0.26 eV).

Fig. 6 Reaction pathway for CO oxidation on Co-^SMoS₂ based on the TER mechanism.

After examining all the four possible mechanisms for CO oxidation on Co-^SMoS₂ monolayer, namely ER, LH, TER and NER mechanisms, we found the LH mechanism to be most favorable pathway for CO oxidation, showing the smallest energy barrier of 0.19 eV, followed by the NER mechanism (0.26 eV). Remarkably, the three pathways, namely NER, LH and TER, all show the RDS energy barrier well below 1.0 eV, indicating that CO oxidation can be facilitated on the surface of Co-^SMoS₂ at low temperatures. The performance of Co/MoS₂ monolayer is finally examined as a comparison of Co-^SMoS₂. It can be found that the LH mechanism (Fig. S7, ESI†) is also the most preferred pathway for CO oxidation on Co/MoS₂ by showing an energy barrier of 0.62 eV. These results are highly consistent with those of the previous study (LH, 0.57 eV)⁴³ demonstrating the reliability of our calculations. As a result, our investigations indicate that Co-^SMoS₂ is a promising catalyst for CO oxidation and shows better stability and catalytic activity than that of Co/MoS₂.

The preferred LH mechanism for CO oxidation on Co-^SMoS₂ can be understood by analyzing the PDOS of the IS1 and TS1 structures along the LH mechanism on Co-^SMoS₂ (Fig. S8, ESI†). As shown in Fig. S8a,† the 2π* orbital of the adsorbed O₂ is partially occupied and Bader charge analysis indicates that there is about 0.36|e| charge transferred from Co to O₂, indicating that the O–O bond is activated by Co of Co-^SMoS₂. In addition, there is a significant overlap between the 5σ* and 2π* states of O₂ and the states of Co atom, demonstrating the strong interaction between O₂ and Co. From IS1 to TS1, there is about 0.03|e| charge transfer from CO to O₂ and the 5σ* state of O₂ partially shifts closer to the Fermi level (Fig. S8b†). These results indicate that the co-adsorption of CO can further weaken the O1–O2 bond O₂. It can also be found that the 5σ and 1π states of CO and O₂ are broadened, indicating the formation of OCOO (IM). The above results indicate that the LH mechanism is favored by the activation of the O1–O2 bond by the embedded Co atom and the co-adsorbed CO.

To better evaluate the catalytic activity of Co-^SMoS₂ for CO oxidation, we compared the performance of Co-^SMoS₂ with that of other noble metal-based SACs and the energy barriers of the RDS over various catalysts are presented in Table 1. The energy barrier for CO oxidation on Co-^SMoS₂ can be found to be lower than that of other reported noble metal-based SACs, indicating that Co-^SMoS₂ is a highly promising SAC candidate for CO oxidation by showing higher catalytic activity and lower cost than many noble metal-based SACs.

Table 1 The energy barriers for the RDS of the most favorable mechanism of CO oxidation on Co-^SMoS₂ in comparison with that on various noble metal-based SACs

Catalyst	Barrier (eV)	Mechanism	Ref.
Co-^SMoS2	0.19	LH	This work
Co/MoS2	0.57	LH	43
Pt-graphene	0.59	LH	64
Pd-graphene	0.20	LH	65
Au-graphene	0.31	LH	66
Pd-BN nanosheet	0.35	TER	67
Au-BN	0.47	LH	34
Pd-BN	0.39	TER	68
Pt-GO	0.76	LH	69
Ir-graphdiyne	0.37	NER	70

---

4. Conclusions

As a summary, by performing DFT calculations, we have studied the possibility of using Co-embedded MoS₂ monolayers as SACs for CO oxidation. The results show that Co atoms embedded at the S vacancies of MoS₂ can form stable SACs without aggregating due to the strong hybridization between Co and Mo atoms, which is more robust than the Co doped defect-free MoS₂ surface. The adsorption manner of CO and O₂ molecules on Co-^SMoS₂ can decide the reaction pathway of CO oxidation and the co-adsorption of CO + O₂ and two CO are both favorable at the Co site of Co-^SMoS₂. The four possible reaction pathways, including ER, LH, TER and NER mechanisms for CO oxidation on Co-^SMoS₂ were discussed and the results show that the oxidation of CO can be facilitated through the LH mechanism by overcoming a small energy barrier of 0.19 eV, which is superior to many noble-metal based SACs. By exhibiting good stability, high catalytic activity, and low-cost characteristics, Co-^SMoS₂ is predicted to serve as a promising non-noble metal-based SAC candidate for efficient catalytic CO oxidation. Since Co-^SMoS₂ have been experimentally synthesized, the practice of using this material for CO oxidation can be expected in the near future.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

We appreciate the financial support in China by Natural Science Foundation of Jiangsu Province of China (no. BK20190744) and National Natural Science Foundation of China (no. 21903046, 22222304).

References

1. J. J. Baschuk and X. Li, Int. J. Energy Res., 2001, 25, 695–713.
2. F. Mariño, C. Descorme and D. Duprez, Appl. Catal., B, 2004, 54, 59–66.
3. L. Birry, J. H. Zagal and J.-P. Dodelet, Electrochem. Commun., 2010, 12, 628–631.
4. L. Cao, W. Liu, Q. Luo, R. Yin, B. Wang, J. Weissenrieder, M. Soldemo, H. Yan, Y. Lin, Z. Sun, C. Ma, W. Zhang, S. Chen, H. Wang, Q. Guan, T. Yao, S. Wei, J. Yang and J. Lu, Nature, 2019, 565, 631–635.
5. H. Zhang, S. Fang and Y. H. Hu, Catal. Rev.: Sci. Eng., 2022, 64, 491–532.
6. B. Qiao, J. Liu, Y.-G. Wang, Q. Lin, X. Liu, A. Wang, J. Li, T. Zhang and J. Liu, ACS Catal., 2015, 5, 6249–6254.
7. H. Su and Y. H. Hu, Chem. Eng. J., 2020, 402, 126235.
8. T. Bunluesin, H. Cordatos and R. J. Gorte, J. Catal., 1995, 157, 222–226.
9. A. Alavi, P. Hu, T. Deutsch, P. L. Silvestrelli and J. Hutter, Phys. Rev. Lett., 1998, 80, 3650–3653.
10. A. Beniya, S. Higashi, N. Ohba, R. Jinnouchi, H. Hirata and Y. Watanabe, Nat. Commun., 2020, 11, 1888.
11. Y. G. Wang, Y. Yoon, V. A. Glezakou, J. Li and R. Rousseau, J. Am. Chem. Soc., 2013, 135, 10673–10683.
12. L. Li, Y. Gao, H. Li, Y. Zhao, Y. Pei, Z. Chen and X. C. Zeng, J. Am. Chem. Soc., 2013, 135, 19336–19346.
13. C. J. Zhang and P. Hu, J. Am. Chem. Soc., 2001, 123, 1166–1172.
14. F. C. H. Lim, J. Zhang, H. Jin, M. B. Sullivan and P. Wu, Appl. Catal., A, 2013, 451, 79–85.
15. H. Öström, H. Öberg, H. Xin, J. LaRue, M. Beye, M. Dell’Angela, J. Gladh, M. L. Ng, J. A. Sellberg, S. Kaya, G. Mercurio, D. Nordlund, M. Hantschmann, F. Hieke, D. Kühn, W. F. Schlotter, G. L. Dakovski, J. J. Turner, M. P. Minitti, A. Mitra, S. P. Moeller, A. Föhlisch, M. Wolf, W. Wurth, M. Persson, J. K. Nørskov, F. Abild-Pedersen, H. Ogasawara, L. G. M. Pettersson and A. Nilsson, Science, 2015, 347, 978–982.
16. O. Khalid, T. Weber, G. Drazic, I. Djerdj and H. Over, J. Phys. Chem. C, 2020, 124, 18670–18683.
17. B. Qiao, A. Wang, X. Yang, L. F. Allard, Z. Jiang, Y. Cui, J. Liu, J. Li and T. Zhang, Nat. Chem., 2011, 3, 634–641.
18. L. Nie, D. Mei, H. Xiong, B. Peng, Z. Ren, X. I. P. Hernandez, A. DeLaRiva, M. Wang, M. H. Engelhard, L. Kovarik, A. K. Datye and Y. Wang, Science, 2017, 358, 1419–1423.
19. T. Kropp and M. Mavrikakis, ACS Catal., 2019, 9, 6864–6868.
20. J. Hu, W. Liu, C. Xin, J. Guo, X. Cheng, J. Wei, C. Hao, G. Zhang and Y. Shi, J. Mater. Chem. A, 2021, 9, 24803–24829.
21. Z. Zhu, H. Yin, Y. Wang, C.-H. Chuang, L. Xing, M. Dong, Y.-R. Lu, G. Casillas-Garcia, Y. Zheng, S. Chen, Y. Dou, P. Liu, Q. Cheng and H. Zhao, Adv. Mater., 2020, 32, 2004670.
22. X. Li, P. Cui, W. Zhong, J. Li, X. Wang, Z. Wang and J. Jiang, ChemComm, 2016, 52, 13233–13236.
23. S. Dou, C.-L. Dong, Z. Hu, Y.-C. Huang, J.-l. Chen, L. Tao, D. Yan, D. Chen, S. Shen, S. Chou and S. Wang, Adv. Funct. Mater., 2017, 27, 1702546.
24. Y. Xue, B. Huang, Y. Yi, Y. Guo, Z. Zuo, Y. Li, Z. Jia, H. Liu and Y. Li, Nat. Commun., 2018, 9, 1460.
25. X.-P. Yin, H.-J. Wang, S.-F. Tang, X.-L. Lu, M. Shu, R. Si and T.-B. Lu, Angew. Chem., Int. Ed., 2018, 57, 9382–9386.
26. X.-F. Yang, A. Wang, B. Qiao, J. Li, J. Liu and T. Zhang, Acc. Chem. Res., 2013, 46, 1740–1748.
27. S. Lin, X. Ye, R. S. Johnson and H. Guo, J. Phys. Chem. C, 2013, 117, 17319–17326.
28. S. Wang, J. Li, Q. Li, X. Bai and J. Wang, Nanoscale, 2020, 12, 364–371.
29. E. W. McFarland and H. Metiu, Chem. Rev., 2013, 113, 4391–4427.
30. J. Liu, ACS Catal., 2017, 7, 34–59.
31. Y.-H. Lu, M. Zhou, C. Zhang and Y. P. Feng, J. Phys. Chem. C, 2009, 113, 20156–20160.
32. E. H. Song, Z. Wen and Q. Jiang, J. Phys. Chem. C, 2011, 115, 3678–3683.
33. Y. Tang, W. Chen, H. Chai, G. Zhao, Y. Li, D. Ma and X. Dai, J. Phys. Chem. C, 2019, 123, 10926–10939.
34. K. Mao, L. Li, W. Zhang, Y. Pei, X. C. Zeng, X. Wu and J. Yang, Sci. Rep., 2014, 4, 5441.
35. S. Manzeli, D. Ovchinnikov, D. Pasquier, O. V. Yazyev and A. Kis, Nat. Rev. Mater., 2017, 2, 17033.
36. K. F. Mak, C. Lee, J. Hone, J. Shan and T. F. Heinz, Phys. Rev. Lett., 2010, 105, 136805.
37. J. Mao, Y. Wang, Z. Zheng and D. Deng, Front. Phys., 2018, 13, 138118.
38. Q. H. Wang, K. Kalantar-Zadeh, A. Kis, J. N. Coleman and M. S. Strano, Nat. Nanotechnol., 2012, 7, 699–712.
39. Y. Fan, J. Zhang, Y. Qiu, J. Zhu, Y. Zhang and G. Hu, Comput. Mater. Sci., 2017, 138, 255–266.
40. D. Ma, W. Ju, T. Li, X. Zhang, C. He, B. Ma, Z. Lu and Z. Yang, Appl. Surf. Sci., 2016, 383, 98–105.
41. D. Ma, W. Ju, T. Li, G. Yang, C. He, B. Ma, Y. Tang, Z. Lu and Z. Yang, Appl. Surf. Sci., 2016, 371, 180–188.
42. Y.-Y. Chen, M. Dong, J. Wang and H. Jiao, J. Phys. Chem. C, 2010, 114, 16669–16676.
43. S. Liu and S. Huang, Appl. Surf. Sci., 2017, 425, 478–483.
44. D. Ma, Y. Tang, G. Yang, J. Zeng, C. He and Z. Lu, Appl. Surf. Sci., 2015, 328, 71–77.
45. A. Sharma, A. Srivastava, M. Husain and M. S. Khan, J. Mater. Sci., 2018, 53, 9578–9588.
46. Q. Ma, P. M. Odenthal, J. Mann, D. Le, C. S. Wang, Y. Zhu, T. Chen, D. Sun, K. Yamaguchi, T. Tran, M. Wurch, J. L. McKinley, J. Wyrick, K. Magnone, T. F. Heinz, T. S. Rahman, R. Kawakami and L. Bartels, J. Phys.: Condens. Matter, 2013, 25, 252201.
47. H.-P. Komsa, J. Kotakoski, S. Kurasch, O. Lehtinen, U. Kaiser and A. V. Krasheninnikov, Phys. Rev. Lett., 2012, 109, 035503.
48. H.-P. Komsa, S. Kurasch, O. Lehtinen, U. Kaiser and A. V. Krasheninnikov, Phys. Rev. B: Condens. Matter Mater. Phys., 2013, 88, 035301.
49. G. Liu, A. W. Robertson, M. M.-J. Li, W. C. H. Kuo, M. T. Darby, M. H. Muhieddine, Y.-C. Lin, K. Suenaga, M. Stamatakis, J. H. Warner and S. C. E. Tsang, Nat. Chem., 2017, 9, 810–816.
50. T. Jitwatanasirikul, T. Roongcharoen, C. Chitpakdee, S. Jungsuttiwong, P. Poldorn, K. Takahashi and S. Namuangruk, New J. Chem., 2021, 45, 17407–17417.
51. P. E. Blöchl, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 50, 17953–17979.
52. G. Kresse and D. Joubert, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 59, 1758–1775.
53. G. Kresse and J. Furthmüller, Comput. Mater. Sci., 1996, 6, 15–50.
54. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865–3868.
55. S. Grimme, S. Ehrlich and L. Goerigk, J. Comput. Chem., 2011, 32, 1456–1465.
56. B. Uberuaga and H. Jonsson, J. Chem. Phys., 2000, 113, 9901–9904.
57. G. Henkelman and H. Jónsson, J. Chem. Phys., 1999, 111, 7010–7022.
58. J. Kästner and P. Sherwood, J. Chem. Phys., 2008, 128, 014106.
59. D. Yang, S. J. Sandoval, W. M. R. Divigalpitiya, J. C. Irwin and R. F. Frindt, Phys. Rev. B: Condens. Matter Mater. Phys., 1991, 43, 12053–12056.
60. M. Chen, X.-B. Yang, J. Cui, J.-J. Tang, L.-Y. Gan, M. Zhu and Y.-J. Zhao, Int. J. Hydrogen Energy, 2012, 37, 309–317.
61. G. Henkelman, A. Arnaldsson and H. Jónsson, Comput. Mater. Sci., 2006, 36, 354–360.
62. G. Xu, R. Wang, Z. Lu, D. Ma and Z. Yang, J. Phys. Chem. C, 2018, 122, 23481–23492.
63. F. Li and Z. Chen, Nanoscale, 2018, 10, 15696–15705.
64. Y. Tang, Z. Yang and X. Dai, Phys. Chem. Chem. Phys., 2012, 14, 16566–16572.
65. M. D. Esrafili, P. Nematollahi and R. Nurazar, Superlattices Microstruct., 2016, 92, 60–67.
66. Y.-H. Lu, M. Zhou, C. Zhang and Y. P. Feng, J. Phys. Chem. C, 2009, 113, 20156–20160.
67. M. D. Esrafili and S. Asadollahi, ChemistrySelect, 2018, 3, 9181–9188.
68. Z. Lu, P. Lv, J. Xue, H. Wang, Y. Wang, Y. Huang, C. He, D. Ma and Z. Yang, RSC Adv., 2015, 5, 84381–84388.
69. Y. Tang, X. Dai, Z. Yang, L. Pan, W. Chen, D. Ma and Z. Lu, Phys. Chem. Chem. Phys., 2014, 16, 7887–7895.
70. G. Xu, R. Wang, Y. Ding, Z. Lu, D. Ma and Z. Yang, J. Phys. Chem. C, 2018, 122, 23481–23492.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d2ra06261e

---