Synergistic effect of Ag(111) and transition metal promoters toward optimization of catalytic ethylene epoxidation selectivity

Contents

Structural optimization and surface models Surface charge analysis Charge density difference analysis Adsorption properties of molecular oxygen Adsorption properties of ethylene Reaction mechanism of ethylene epoxidation Catalytic activity and selectivity toward EO product

Structural optimization and surface models

Figure 1 presents the top and side views of optimized structures of promoted Ag catalysts. The Ag (111) surface was selected as the representative surface for the catalyst due to its superior thermodynamic stability and its prevalence in commercial EO catalysts^44,45,46. The Ag (111) surface was modeled using a 4 × 4 supercell consisting of three atomic layers of Ag. The largest species examined have an effective molecular diameter of 3.53 Å, with a center-to-center distance of 20.93 Å between two adjacent unit cells. Since every single elementary step in the reaction occurs independently on the surface of the supercell, the analysis of the reaction path itself needs to be focused entirely on those species participating in the particular action, typically two atoms or molecules. Hence, the supercell size is large enough to represent all possible relative positions among these species, with negligible impacts on the next nearest neighbor (NNN). The surface adsorbates and the upper two metal layers were allowed to relax to optimize the surface structure. For simplicity, the lower layer was constrained to the positions of bulk Ag atoms.

Before we can look at how different promoters work together to improve the selectivity of EO during ethylene epoxidation, we need to figure out where on the Ag (111) surface each promoter is most thermodynamically optimal. In order to obtain a structure that reflects the fractional occupancy, we used the Mixture Atoms module in the Materials Studio software. By applying the Mixture Atoms module, we can simulate cases where different types of atoms occupy an atomic site. Namely, for a dual-promoter catalyst, the composition of a Cs promoter was altered to 50% Cs and 50% Au. This promoter can either bind to the surface Ag atoms as an adsorbate or become part of the surface of the Ag slab as a dopant.

In Fig. 1, the adsorption sites for the promoters (Cs, Re, Rh, Au, Cu, W, Zn, Mo, and their combination) on Ag (111) surfaces, face-centered cubic (FCC), Bridge (B), and hexagonal-close packed (HCP) are provided. Table S1 presents the adsorption energies for all the concrete promoters. The binding energy calculations reveal that the FCC site is the most favorable for the promoters to bind. The chemical bonds between the Ag and the promoter atom are formed by interactions between the d-states of the prompter and the in-plane d-states of the Ag atoms. In the optimized Ag/Cs-Au nanosheet, the metallic bonds formed between Cs-Au and three neighboring Ag atoms have a bond length of about 2.75 Å, which is much shorter than the Ag–Ag bonds of 2.91 Å in the pristine Ag (111) nanosheet, as shown in Table S1. The Ag–Ag bond lengths determined in the present study are slightly longer than the value reported by Lee et al. (2.89 Å)⁴⁷ and slightly shorter than that reported by Li et al. (3.10 Å)⁴⁸.

Table S1 collects a summary of the BEs related to the interactions of the promoter with Ag atoms in the models developed. A lower BE value corresponds to a stronger binding affinity of the substrate to the silver. Adding either individual or mixed metal promoter moiety—regardless of their specific nature—generally results in increased substrate BEs for most systems compared to pristine Ag. Remarkably, TM dopants exhibit strong binding properties, which might account for the enhanced catalytic activity revealed experimentally^31,49.

Surface charge analysis

It is believed that nucleophilic and electrophilic oxygen atoms are responsible for both the nonselective and the selective processes^50,51. Our study begins with a Hirshfeld charge analysis of oxygen at the atomic level. In the charge analysis, the oxygen atom has a significant negative charge (− 0.390 e), whereas the Ag atom has a negative net charge accumulation (− 0.007 e). The adsorbed O atom acquires a substantial number of electrons from the substrate. Consequently, the net charge accumulation on the defective Ag becomes significantly positive, from − 0.007 e to 0.034 e. The initial value, i.e., −0.007 e, indicates that the silver atom featured a slightly negative charge, indicating a small excess of electrons. Also, the shift to 0.034 e indicates the transition of the atom to a positively charged state, which reflects the increased positive nature of the atom’s environment or a net loss of electrons.

The atomic O becomes negatively charged upon the addition of Cs, thus increasing its nucleophilic. Conversely, TMs transform that charge into a more electrophilic behavior. This is consistent with the electron-donating and electron-receiving behaviors of Cs and TMs, respectively. According to Hirshfeld charge analysis, the Re atom contributes approximately 0.26 electrons to the surface of the Ag/Cs-Re particles. In addition, due to a large amount of charge transfer, strong interactions occur between the positively charged Re and its neighbor Ag atoms. This interaction is also supported by the calculated high adsorption energy of the Cs-Re atom on the defective Ag (4.64 eV). The Cs-Re atom has a Hirshfeld charge density of 0.68 au, with the most unpaired electrons among all considered atoms. Based on the large positive atomic charge calculated, Ag/Cs-Re becomes the most active nucleophilic or radical attack site on the surface.

The Hirshfeld charge of Ag exhibits a similar pattern to that of Ag/Cs-Re, with Ag becoming more positively charged with introducing transition metal(s) and more negatively charged with adding Cs. This similarity implies that charge transfer from promoters to the catalyst surface is responsible for the Hirshfeld charge trends in atomic O and Ag. In transitional metals, electron density is accepted from silver, which subsequently interacts with atomic oxygen to increase the metal’s electrophilicity. Similar coverage-dependent effects have been observed in the adsorption of O on the surfaces of various TM(s) and alloys^48,52. Furthermore, donating promoter charges facilitates filling the O’s p-bands and s-bands, expanding band centers. As a result, given that the p-band center of O has a broader range than the d-band center of Ag, the (O p-Ag d) band center is expected to be influenced (see Figure S4). The ionic nature of the Ag/O bond suggests that an increased O charge density improves the bonding strength of O to the surface⁵³. In addition, nucleophilic oxygen species show considerable binding affinity to mixtures of Cs with transition metals. These results complement previous DFT studies for several such mixtures^19,53 and experiments²⁷. This evidence suggests that while Cs strengthens the affinity of O, transition metals (TMSs) tend to weaken it.

Charge density difference analysis

A charge density difference (CDD) plot was used to analyze the redistribution of electron density. Figure 2 shows the CDD of the developed catalysts, calculated by the following equation:

$$\Delta \rho ={\rho }_{Cat}-{\rho }_{Ag}-{\rho }_{Promoter}-{\rho }_{O}$$

(5)

where ${\rho }_{Cat}$, ${\rho }_{Ag}$, ${\rho }_{Promoter}$, and ${\rho }_{O}$ are the charge densities of the simulated catalyst, pristine Ag (111) slab, promoter(s) atom, and pristine oxygen adsorbed systems, respectively^54,55. A cutting slice, defined by the distance between the promoter atom and the nearest O atom, was oriented perpendicular to the Ag plane to illustrate the CCD. In the resulting visualization, charge accumulation and depletion by the promoters, O, and Ag atoms are presented in red and blue colors, respectively. Charge depletion is observed in most promoter atoms, except for Cs-Zn and Cs-Mo, while electron accumulation is observed repeatedly in the O atoms (Figs. 2a–o). This electron density accumulation between the O atom and the surface indicates that oxygen is polarized by the positively charged Ag atom on the surface. As previously reported^56,57,58, the electrophilic nature of the O atom has been shown to improve EO selectivity.

Furthermore, the increased electron density surrounding the O atom results in a reduced transfer of electrons from the catalyst to the O atom, promoting the formation of electrophilic O atoms. Conversely, TM atoms enhance the electrophilic nature of O by reducing the accumulation of electron density from adjacent O atoms. This process is beneficial for producing more selective sites for ethylene epoxidation since the Cs atoms on the Ag surface increase the electron density accumulation. In addition, as shown in Fig. 2, the electronic charge density in pristine Ag (111) is spherically symmetrical around the O atoms, while that in promoted-Ag (111) catalysts is significantly asymmetrical. This kind of charge distribution is caused by factors such as valence electron delocalization or directional bonding, which substantially affects the deformation behavior of the catalysts^59,60.

The findings confirm the previous hypothesis that selectivity gain is strongly connected to the promoter electric effects. In fact, previous works^22,61, have reported that the promoters generate an electric field that modifies the energy of intermediates and transition states. Thus, the mixed TMs promoters, especially Cs₂₂Cu₂₄Re₂₄Au₃₀ catalyst, would potentially show higher efficiency in selectivity and activity of the ethylene epoxidation reaction. This is due to the enhancement in electrophilicity of the atomic oxygen.

Adsorption properties of molecular oxygen

The adsorbed O atom is critical in increasing EO selectivity in the ethylene epoxidation reaction process⁶². The active species was previously thought to be the O₂ molecule²⁵. The first stage is to characterize the adsorption of O₂ molecules and O atoms on Ag surfaces, whether pristine or promoted with mixed metal. It is important to note that the ethylene epoxidation process may include a stage where molecular oxygen dissociates on the catalysts. Therefore, we investigated the adsorption of molecular oxygen as a potential starting point for the decomposition of O₂. Our study used the adsorption locator module in Materials Studio to explore various molecular oxygen adsorption configurations on the surface of the catalysts. The module suggested several configurations based on their adsorption energies. Using a systematic evaluation of multiple configurations, the most favorable configuration (i.e., the most negative adsorption value) was assigned the one with the highest adsorption energy. Following a series of studies, Table 1 summarizes the most favorable O₂ adsorption configurations across various catalysts. Previous computational studies estimate that the energy released by O₂ adsorption on pristine Ag varies between −0.243 and −0.70 eV, depending on the computational techniques, high and low oxygen coverage, and crystal size^63,64,65,66. Specifically, Zhu et al.⁶³ showed an adsorption energy of −0.57 eV using DFT calculations employed by PWSCF (Plane-Wave Self-Consistent Field) code in the Quantum ESPRESSO package. However, lower adsorption energy (−0.244 eV) using the DMOL³ code in the present study might be due to using different computational codes in calculating O₂ adsorption energies. Notably, Yu et al.⁶⁷ reported an adsorption energy of -0.243 eV for O₂ on Ag (111) using the projector augmented wave (PAW) code, which closely aligns with our calculated value.

Table 1 Calculated adsorption energies of molecular oxygen (in eV), Ag-O₂ bond length (in Å), and local configuration of the molecular oxygen adsorption on the promoted and unpromoted surfaces.

Because of the weak interaction of molecular oxygen with promoter atoms, O₂ adsorption only slightly increases in mixed metal-promoted catalysts. The larger Ag–O₂ distance in the Cs–Ag model confirms that the Cs oxyanion plays an insignificant promotional role in dioxygen activation. Notably, the Cs atom is positioned outside the plane formed by the Ag atoms in all Cs-containing models. This arrangement is significant since it influences the interaction between the Ag and Cs atoms and their adsorption properties. This out-of-plane positioning may be influenced by the large size of the Cs atoms, although it is also important to consider the effect of the restructured surface. As depicted in Fig. 1, O atoms can occupy three possible adsorption sites on both pristine and mixed metal-promoted Ag (111) surfaces: FCC, B, and HCP. Among these, the FCC site is confirmed to be the most stable atom for O adsorption. The current study shows that Cs cannot form a covalent bond with the O atom on the Cs-Ag catalyst. The adsorption energies of the atomic oxygen at the most favorable site (i.e., FCC) on the pristine and mixed metal-promoted Ag are presented in Table 1. For instance, the calculated adsorption energy of the O atom on the pristine Ag (111) surface is -5.05 eV. On the other hand, Ren et al. reported the binding energy of the O atom on the same Ag (111) surface as –5.92 eV²³. The differences in the observed binding energies can be attributed to the smaller size of the silver clusters, known to show unique properties compared to larger particles or surfaces, including increased reactivity⁶⁸.

The mixed metal-promoted models show higher binding energies for most of the interactions investigated^69,70,71,72. Notably, when Cs and Rh moieties are combined, the binding energies of atomic oxygen are significantly increased compared to the pristine Ag model. In the Ag/Cs-Rh catalyst, the energy values associated with substrate bonding with the surface are doomed “optimal,” aligning with the experimentally observed superior performance of Ag catalysts co-promoted by Rh and Cs when compared to those promoted individually by either Rh or Cs. This observation indicates the synergistic effect of the Rh and Cs promoters. Furthermore, this enhancement suggests an increase in the strength of facet-dependent adsorption. It has been demonstrated that the adsorption energy of the adsorbate on these catalysts is proportional to the d-band center (εd) of the occupied d-d-band (see Figure S4)^73,74.

The results for O₂ adsorption are also in acceptable agreement with those reported by Zhu et al.⁶³, where the atomic oxygen binding energy results are similar to those obtained by Ren et al.²³. The Ag–O bond length for O atom binding is consistent with the value calculated by Zhu et al.⁶³ for Ag (111) FCC sites. Notably, the Ag–O bond values indicate that the strength of the O atom is significantly greater on mixed metal-promoted Ag catalysts than on pristine Ag, as shown in Table S1. According to Table 1, the adsorption energy of the O atom on the Cs₂₅Cu₂₅Re₂₅Au₂₅-promoted Ag (111) surface increases to −7.77 eV, compared to −5.05 eV on the pristine Ag (111) surface.

Adsorption properties of ethylene

The adsorption of C₂H₄ and O atoms on the surfaces of the developed catalysts has been calculated to gain a deeper understanding of the reaction mechanism on both pristine and mixed metal-promoted Ag surfaces. Table 2 summarizes the optimized stable structures and adsorption energies of ethylene on these Ag (111) surfaces. Tables S2, S3, and S5 indicate the presence of atomic oxygen species on the entire surface, at different positions of Ag atoms, and coordinates of O-adsorbed on the Ag/Cs-Au surface simultaneously. In addition, the atomic or lattice oxygen occupying the FCC positions acts as the active site for ethylene adsorption. The optimized geometrical parameters of ethylene are compared quite well with the theoretical values reported by Bocquet and Loffreda^75,76. For example, the C–H and C–C bond lengths were found to be 1.09 Å and 1.34 Å, respectively, while the angles H–C–C and H–C–H were to be 121° and 117°. In our model, the calculated values for these parameters are 121.9° and 116.3°.

Table 2 Calculated adsorption energies of molecular oxygen (in eV), O-C bond length (in Å), and local configuration of the ethylene adsorption on the promoted and unpromoted surfaces.

As shown in Table 2, atomic oxygen species are located on top of coordinatively unsaturated (CU) Ag atoms. All Ag_cu sites can be fully covered with atomic oxygen by exposing the Ag (111) surface to multiple Langmuirs of molecular oxygen at various temperatures and oxygen pressures^77,78. However, experimental evidence demonstrates that ethylene adsorption requires an empty Ag_cu site. Without an empty Ag_cu site, ethylene cannot bind atop the oxygen atoms⁷⁸. To evaluate the current state of the ethylene epoxidation reaction, an ethylene gas molecule was positioned over the oxygen atoms to assess binding. According to calculations of these configurations, the ethylene molecule adsorbs at oxygen atom sites. The binding distance of ethylene to the Ag atom on the surface of a pristine Ag catalyst is measured at 3.71 Å.

Furthermore, ethylene aligns nearly parallel to the nanosheet surface, attributed to a favorable orbital interaction between its d orbital and the d orbital of the nanosheet surface. C₂H₄ has an adsorption energy of −0.082 eV on pristine Ag surfaces (111). According to Gao et al.⁷⁹, ethylene adsorbed at a rate of −0.119 eV onto Ag sites. The differences in binding energies can be explained by the small size of the silver clusters, which are known to possess unique properties compared to larger particles or surfaces, including enhanced reactivity⁶⁸. As shown in Table 2, the adsorption of C₂H₄ on mixed metal-promoted Ag (111) surfaces shows greater stability than on pristine Ag (111) surfaces. Figure 3 illustrates the adsorption strength of C₂H₄ on Ag (111) surfaces, indicating that the promoters exert a facet-dependent effect on the adsorption strength of the ethylene.

Reaction mechanism of ethylene epoxidation

The conversion of C₂H₄ to EO is depicted in Fig. 4 as a two-dimensional structural process. The minimum energy pathway for the reaction is also shown in Fig. 5, which includes the optimized structures of the reactant, transition state (TS), product structures, and relative energies. The reaction consists of ten fundamental steps, summarized and organized in Table 3. Firstly, ethylene and molecular oxygen are weakly adsorbed onto the silver surface (steps 1 and 2). The dissociative adsorption of the adsorbed molecular or gaseous oxygen generates atomic oxygen on the surface (steps 3* and 3). There are two potential scenarios in step 4: co-adsorption of oxygen and ethylene (step 4*) or ethylene adsorption occurring through an Eley–Rideal mechanism (step 4). Following the OMC intermediate, the reaction continues through two competing pathways previously observed on other surfaces^80,81.

Table 3 List of elementary steps for partial and complete oxidation of ethylene on the developed catalysts¹².

The first pathway builds up an epoxy ring on the Ag surface by breaking the C = C bond to form an EO molecule (step 5). This elementary reaction is exothermic, while its reaction energy of −0.63 eV indicates it may occur under typical conditions. The conversion of EO to AA through isomerization is illustrated in Fig. 4 (Step 6). Because of the strong chemical interaction between the C and the displaced H atoms, this reaction is more exothermic than the previous step, but it has a higher activation energy barrier of 1.06 eV. When assessing the degree of rate control, these processes are identified as rate-determining and selectivity-controlling⁸². The subsequent desorption of EO and AA molecules into the gas phase after OMC isomerization does not require a transition state and involves negligible reaction energy (steps 7 and 8). The competitive oxidation reaction to AA immediately turns into H₂O, and CO₂ is thermodynamically favorable because $\Delta G=-194$ kJ.mol^-1 ⁶¹. In other words, the energy barrier for transforming AA to CO₂ and H₂O is much lower than that for OMC to AA. Therefore, it is reasonable to conclude that the complete combustion reaction is not a rate-determinate step in the EO reaction.

According to Jankowiak and Barteau⁸³, the activation energy barrier for ethylene epoxidation is 0.58 eV, which is in acceptable agreement with our 0.66 eV on pristine Ag (111), considering the bulk metallic catalyst is primarily composed of Ag (111). Grant and Lambert⁸⁴ represented that the activation energy barriers for the mixture of C₂H₄ and O₂ in a 1:1 ratio range between 0.45–0.53 eV and 0.52–0.63 eV, respectively, which are within our simulated results. Previous studies have also indicated that activation energy barriers can range between 0.4 and 1.1 eV, depending on experimental conditions and laboratory setups^85,86.

This work does not include a complete oxide phase because, in general, metal oxides have a lower affinity for oxygen than metals, resulting in dramatically different scaling relationships^87,88. Therefore, the present study concentrates on the general features of ethylene oxidation for small oxygen coverages. Previous computations of the O-Ag phase diagram as a function of pressure and temperature⁸⁹ revealed that under typical experimental conditions for ethylene epoxidation—characterized by low O₂ pressure (0.1 bar) and high temperatures (230 °C)— a low O-coverage on Ag (111) is thermodynamically beneficial. Furthermore, previous studies on the Ag phase diagram confirm that under the reaction conditions for ethylene epoxidation, the Ag₂O surface is unstable. In fact, the Ag surface partially oxidized during the reaction^45,46,90.

Moreover, prior investigations that employed first principles-aided thermodynamic modeling to study the phase diagrams of the TMs in the oxygen environment (Ag⁸⁹, Au⁹¹, Pt⁹², Cu⁹³, Pd⁹⁴, Ir⁹⁵, Ni⁹⁶, Ru⁹⁷, Rh⁹⁸) indicate that Pt and Au promoters maintain their metallic phase during ethylene epoxidation. In contrast, the promoters Ni, Cu, Ir, Pd, Ru, and Rh fully change to their corresponding oxide states. According to XPS spectra, Diago et al. show that Re and Mo oxidize in a pure oxygen environment, turning into Re₂O₇ and MoO₃ phases, respectively. However, it should be noted that in real conditions of the ethylene epoxidation reaction, the oxygen concentration is about 20%, suggesting that these promoters may not be fully oxidized. Thus, based on the partially oxidized Ag surface, our present models could be relevant to the mechanism of EO production under low-pressure oxygen conditions.

To determine which catalysts are more active toward producing either ethylene oxides (EOs) or acetaldehydes (AAs), investigations into the transition state (TS) features of the two competing processes must be pursued. Figure 5 depicts the TS values for both EO and AA after OMC isomerization. The top and side views of OMC, EO, and AA formation over-developed catalysts are presented simultaneously in Figures S1–S3. Moreover, the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) of specious are shown in Figure S5 and Table S4. A crystal-clear distinction is observed in the two TSs. In EO formation, the TS includes the breaking of the C = C bond to allow the epoxy ring closure. During this process, oxygen shifts beneath the ethylene fragment while the O-metal bond remains intact. On the other hand, the TS for producing AA involves the dissolution of both C-metal and O-metal bonds with the migration of H from one C atom to another, where a C = O bond is formed. These configuration properties at the two TSs are consistent across all TMs and metal oxide surfaces evaluated in this study. Experimental results indicate that the promoted and unpromoted catalysts produce significantly different amounts of EO. For instance, the activation energy barrier for EO formation is lower on an Ag surface that has been promoted with Cs-Re than a pristine Ag surface (see Fig. 5). Consequently, AA generation on a pristine Ag (111) surface is more efficient than on an Ag (111) surface enhanced with Cs-Re.

Catalytic activity and selectivity toward EO product

Figure 5 illustrates the use of the MEP diagram of both promoted and unpromoted Ag (111) surfaces to compute their theoretical selectivity ($\Delta \Delta {E}_{Activation}$) for EO (see Fig. 6). The $\Delta \Delta {E}_{Activation}$ on the promoted Ag (111) surface, as shown in Fig. 6, is lower than that of $\Delta \Delta {E}_{Activation}$ on a pristine Ag (111) surface. Further details are provided in the text. Previous reports showed that EO selectivity is greater on promoted Ag (111) surfaces than on pristine Ag (111) surfaces^99,100. The activation energy barriers of the OMC intermediate to EO are lower on promoted Ag (111) surfaces, indicating that this surface is more prone to isomerization. In contrast, the activation energy barrier for isomerization of OMC intermediate into AA is higher on prompted Ag (111) than a pristine Ag (111) surface. Thus, in terms of selectivity and catalytic activity, the promoted Ag (111) surface outperforms the pristine Ag (111) surface (see Fig. 6). As a result, the selectivity and catalytic activity for EO on the promoted Ag (111) surface are much higher than those on the pristine Ag (111) surface. In particular, the Ag/Cs₂₂Cu₂₄Re₂₄Au₃₀ promoted Ag (111) surface may obtain the best selectivity for EO among various pristine and mixed-promoted Ag (111) surfaces considered. The following order of EO selectivity would be expected between the promoted and unpromoted Ag (111) surfaces:

pristine Ag < Ag/Cs-Mo < Ag/Cs_33.3Au_33.3Cu_33.3 < Ag/Cs₂₅Cu₂₅Au₂₅Re₂₅ < Ag/Cs-Cu < Ag/Cs-Zn < Ag/Cs < Ag/Cs_33.3Re_33.3Au_33.3 < Ag/Cs₂₂Rh₂₄Re₂₄Au₃₀ < Ag/Cs-W < Ag/Cs_33.3Cu_33.3Re_33.3 < Ag/Cs-Rh < Ag/Cs-Au < Ag/Cs-Re < Ag/Cs₂₂Cu₂₄Re₂₄Au₃₀.

Figures 6 and 7 show that adding TM promoters can improve selectivity and activity, consistent with prior studies^8,30,31,101. On an Ag/Cs₂₂Cu₂₄Re₂₄Au₃₀ promoted Ag (111) surface, for instance, $\Delta \Delta {E}_{Activation}$ increased from 1.29 to 2.71 eV compared to pristine Ag. A higher selectivity value indicates that the activation energy for acetaldehyde formation is greater than that for ethylene oxide formation. Consequently, the elevated activation energy (barrier energy) results in a slower formation rate of acetaldehyde compared to ethylene oxide. Compared with a Cs-promoted Ag (111) surface, the activation energy increased by only 0.69 eV, from 1.29 to 1.98 eV. Notably, Cu as a lone promoter may not positively affect EO selectivity. Consequently, Cu plays an important role when combined with other promoters to achieve high EO selectivity. Indeed, Cu improves the EO selectivity when combined with promoters such as Cs, Au, and Re, among others^29,30.

Our findings indicate that the mixed TM-promoted Ag (111) surface exhibits better EO selectivity and activity than the common Cs-promoted Ag surfaces. That would imply mixed TM catalysts can act efficiently to catalyze the ethylene epoxidation reaction, such as Ag with Ag/Cs₂₂Cu₂₄Re₂₄Au₃₀, for better performance of ethylene oxide production.

It was found that highly active sites, such as those present on Ag (111), can significantly influence the overall reaction kinetics⁴⁹. However, practically speaking, Ag catalysts can be affected by various factors, such as additional active sites and reaction conditions. Therefore, while the simulations in this study mainly focus on the intrinsic characteristics of different Ag facets under specific conditions, it is important to highlight that catalytic performance involves a combination of various facets, including Ag (111), Ag (100), Ag (110), as well as vacancies and defects. According to earlier investigations, Ag (111) surfaces show higher thermodynamic stability in comparison with both Ag(100) and Ag (110) surfaces. Such improved stability could slow down deactivation during reactions and longer catalyst lifetimes, rendering Ag (111) a preferred crystalline plane for a selective EO reaction. By contrast, Ag (100) and Ag (110) surfaces were characterized by lower stability and may undergo deactivation/reconstruction under certain reaction conditions, adversely affecting their efficiency. Indeed, while Ag (110) surfaces may show distinct reactivity patterns and high selectivity, their overall activity (e.g., longevity and stability) may not compete with that of Ag (111)⁸².

Moreover, the catalytic activity on the Ag (111) surfaces, especially when modified by the promoters, can be superior^32,33. This is also due to the incorporation of promoters leading to improved adsorption of intermediates, which results in enhanced reaction rates.

Source link

Synergistic effect of Ag(111) and transition metal promoters toward optimization of catalytic ethylene epoxidation selectivity