2.1. Cu4Ag1 and Cu1Ag4 Bimetallic Clusters
Figure 2a,b depict five optimized Cu4Ag1 trapezoidal bimetallic clusters and five optimized Cu1Ag4 trapezoidal bimetallic clusters using GGA/PBE. For both cases, the ground states of these clusters are two-dimensional trapezoidal configurations, consistent with previous studies [37]. For Cu4Ag1 clusters, the most stable bimetallic cluster is observed when the Ag atom is positioned at the bottom corners of the cluster (structures 1 and 5 in Figure 2a). This result implies that the bottom corner sites promote a more favorable electronic environment for the Ag atom, possibly owing to enhanced bonding interactions with the neighboring Cu atoms. The metastable state takes place when the Ag atom is located at the top corners of the cluster (structures 2 and 3 in Figure 2a), with an energy difference of approximately 0.07 eV compared to the most stable cluster. This metastable state indicates that the top corner positions are less favorable but still within a relatively low energy range, signifying potential flexibility in the structural variation of the clusters. The least stable structure is recognized when the Ag atom occupies the center of the cluster (structures 4 in Figure 2a), revealing an energy difference of 0.23 eV. Figure 2a also confirms that changing a Cu atom with an Ag atom at various sites within the cluster induces different degrees of deformation. This deformation is ascribed to changes in bond length, aligning with prior research [38]. Moreover, the calculated average bond length of Cu-Cu is 2.38 Å and of Cu-Ag is 2.51 Å, which is close to the bulk value of 2.53 Å [39].
For Cu1Ag4 clusters, the most favorable configuration for the bimetallic cluster is achieved when the Cu atom is situated at the center of the cluster, as observed in structure 1 of Figure 2b. This stability is consistent with previous studies [37,40,41]. The observed stability implies that the central sites provide a more favorable electrical environment for the Cu atom, possibly due to increased bonding interactions with the surrounding Ag atoms. A metastable condition occurs when the Cu atom is positioned at the upper corners of the cluster (structures 4 and 5 in Figure 2b), with an energy difference of about 0.07 eV compared to the most stable cluster. This indicates that although the top corner positions are not immensely favorable, they nevertheless fall within a low energy range, demonstrating a certain level of structural adaptability. The most unstable cluster is noted when the Cu atom is positioned in the cluster’s bottom corners (structures 2 and 3 in Figure 2b), inducing an energy difference of 0.21 eV. Moreover, Figure 2b reveals that replacing an Ag atom with a Cu atom at different sites of the cluster leads to different levels of distortion. The distortions observed are attributed to variations in the length of chemical bonds. In addition, the average bond lengths computed for Ag-Ag and Cu-Ag are 2.66 Å and 2.55 Å, respectively.
To clarify the charge transfer mechanism and electronic properties of the lowest-energy configuration of Cu4Ag1 and Cu1Ag4 clusters (see structures 1 in Figure 2a,b), we applied the Bader charge analysis technique, calculated the pDOS, and presented the CDD, as illustrated in Figure 3. Bader charge calculations demonstrate that for Cu4Ag1 cluster, the Ag atom gains ~0.18 e− from the Cu atoms, while for Cu1Ag4 cluster, the Ag atoms gains ~0.22 e− from the Cu atom, a phenomenon clearly proven in the CDD plot. The driving force behind this charge transfer is the difference in Pauling electronegativity between Ag (1.93) and Cu (1.90) [39]. Though the variation in electronegativity is relatively small, it is sufficient to induce an obvious electron transfer, revealing the sensitivity of these clusters to subtle electronic changes. The pDOS calculations further complement these analyses by illustrating the electronic states contributing to the bonding and stability of the cluster. For the Cu4Ag1 cluster, it is evidently observed that the states range between −3.3 eV and −0.8 eV and are dominated mainly by the d orbitals of Cu atoms, with the insignificant contribution of the d orbitals of the Ag atom at around −2.7 eV. The contribution of d orbitals of the Ag atom is more pronounced in the energy range between −4.8 eV and −3.35 eV, with a minor hybridization of the d orbitals of Cu atoms at nearly −3.9 eV.
For the Cu1Ag4 cluster, it is evident that the states within the energy range of −2 eV to −1.3 eV are mainly dominated by the d orbitals of Cu atoms, with a slight contribution from the d orbitals of the Ag atom at around −1.6 eV. In the energy range of −5.2 eV to −2.5 eV, the d orbitals of the Ag atom are more explicit, displaying minor hybridization with the d and s orbitals of the Cu atom near −3.8 eV and −2.7 eV, respectively. The incorporation of Cu into the Ag cluster leads to an improved occupation of the s orbitals of the Ag atoms, as clearly apparent by the overlapped state at −1 eV in Figure 2b. This occurrence likely implies that Cu doping facilitates the d-electrons transfer to the Ag atoms within the Cu1Ag4 cluster. This electron transfer was also found in the Ag12Cu cluster [42].
For both clusters, doping an Ag atom in the Cu cluster and vice versa modifies the electronic structure, as evidenced by the changes in the DOS near the Fermi level (see Figure 3b). In particular, there is obvious hybridization between the d orbitals and sp orbitals near the Fermi level in the Cu atom and the p orbitals in the Ag atoms (see Figure 3a). Similarly, the Ag atoms display boosted overlap of their s orbitals with the sp hybridized orbitals of the Cu atom near the Fermi level (see Figure 3b). This leads to electron transfer from the Cu atom to the Ag atoms in both Cu4Ag1 and Cu1Ag4 clusters, resulting in a substantial splitting between the spin-up and spin-down DOS, which is consistent with previous reports [43]. This hybridization can improve the catalytic properties of the clusters by creating active sites with altered electronic environments, which are necessary for facilitating various chemical reactions.
2.2. Cu3Ag2 and Cu2Ag3 Bimetallic Clusters
Figure 4a,b illustrate ten optimized trapezoidal bimetallic clusters of Cu3Ag2 and Cu2Ag3, respectively, utilizing GGA/PBE methodology. In both systems, the clusters’ ground states reveal two-dimensional trapezoidal structures, corroborating the results of earlier research [37,40]. For Cu3Ag2 clusters, the most stable cluster is achieved when two Ag atoms are located at the bottom corners of the cluster (structures 1 in Figure 4a). This stability possibly arises from the bottom corner sites providing a more favorable electronic environment for the Ag atoms, potentially due to improved bonding interactions with the neighboring Cu atoms. A metastable state is noted for structures 3, 4, 5, 9, and 10 (see Figure 4a), with an energy difference of around 0.14 eV compared to the most stable cluster. The least stable structure arises for structures 2, 6, 7, and 8 (see Figure 4a), with an energy difference of 0.27 eV. Figure 4a further demonstrates that substituting of tow Cu atoms with Ag atoms at different sites within the cluster results in varying degrees of deformation, ascribed to alterations in bond lengths. Moreover, the calculated average bond lengths of the most stable cluster are 2.40 Å for Cu-Cu and 2.52 Å for Cu-Ag.
For Cu2Ag3 clusters, the most favorable arrangement is obtained when two Cu atoms are positioned within the center and top corners of the cluster, as depicted in structures 1 and 2 of Figure 4b. The observed stability implies that these sites offer a more favorable electronic environment for the Cu atom, likely due to enhanced bonding interactions with the surrounding Ag atoms. A metastable state is found for structures 4, 5, 6, 7, 8, 9, and 10 (see Figure 4b), with an energy difference of nearly 0.18 eV compared to the most stable cluster. This signifies that, while the bottom corner positions are less favorable, they remain within a low energy range, showing some degree of structural adaptability. The least stable cluster is observed when the Cu atoms are located at the bottom corners of the cluster (structure 3 in Figure 4b), resulting in an energy difference of about 0.32 eV. Moreover, Figure 4b shows that replacing Ag atoms with Cu atoms at various sites of the cluster induces different levels of distortion. These distortions are caused by changes in bond lengths. Furthermore, the computed average bond lengths for the lowest energy state are 2.62 Å, 2.42 Å, and 2.53 Å for Ag-Ag, Cu-Cu, and Cu-Ag, respectively.
To describe the charge transfer mechanism and the electronic properties of the lowest-energy structures of Cu3Ag2 and Cu2Ag3 clusters (see structures 1 in Figure 4a,b) we employed Bader charge analysis, calculated the pDOS, and illustrated the CDD, as presented in Figure 5. The Bader charge calculations indicate that in the Cu3Ag2 cluster, the Ag atoms gain approximately 0.30 e− from the Cu atoms, whereas in the Cu2Ag3 cluster, the Ag atoms gain about 0.41 e− from the Cu atoms. This phenomenon is clearly observed in the charge density difference plots seen in Figure 5a,b. The pDOS calculations offer further insights by identifying the electronic states that contribute to the bonding and stability of the clusters. In the Cu3Ag2 cluster, the energy range from −2.7 eV to −0.7 eV is predominantly influenced by the d orbitals of Cu atoms, with a modest contribution from the d states of the Ag atom around −2.1 eV. The Ag atom’s d orbitals are particularly prominent in the energy range from −4.8 eV to −3.1 eV, with a slight hybridization of the Cu atoms’ d orbitals.
In terms of the Cu3Ag2 cluster, the energy range from −2.5 eV to −0.8 eV is mainly dominated by the d orbitals of Cu atoms, with a small contribution from the d states of the Ag atom around −2.0 eV. The d states of the Ag atom become more prominent in the energy range from −5.0 eV to −2.5 eV, indicating modest hybridization with the d and s orbitals of the Cu atom around −4.0 eV and −2.7 eV, respectively. The incorporation of Cu into the Ag cluster results in a greater filling of the s orbitals of Ag atoms, as evidenced by the overlapping state at −1.1 eV in Figure 5b. This suggests that the addition of Cu facilitates the transfer of d-electrons to the Ag atoms in the Cu2Ag3 cluster. The electronic structure of both clusters is influenced by the doping of Ag and Cu atoms in the Cu cluster and vice versa. This effect is evident from the variations in the DOS near the Fermi level. Specifically, there is clear hybridization between the d states and sp levels near the Fermi level within the Cu atom, as well as the p orbitals within the Ag atoms (see Figure 5a). Similarly, the Ag atoms, which have neighboring atoms, demonstrate pronounced overlap of their s orbitals with the sp hybridized orbitals of the Cu atom around the Fermi level (see Figure 5b). This interaction causes the transfer of charge from the Cu atom to the Ag atoms in both Cu4Ag1 and Cu1Ag4 clusters, resulting in a significant separation between the spin-up and spin-down DOS. This observation is consistent with previous studies [43]. The hybridization can enhance the catalytic properties of the clusters by producing active sites with modified electronic environments, which are essential for facilitating several chemical reactions.
2.3. Adsorption of CO2 on Cu4Ag1 and Cu1Ag4 Clusters
Bimetallic clusters have gathered substantial interest due to their remarkable catalytic activity and reduced susceptibility to CO2 poisoning [44]. The primary aim of this research is to understand the reactivity of CO2 gas molecules on CumAgn bimetallic clusters. Therefore, we explore the influence of compositional changes on the reactivity of CO2 molecules across bimetallic clusters of CumAgn. Before adsorbing the CO2 molecule onto the CumAgn bimetallic clusters, we optimized the isolated CO2 molecule using GGA/DFT (see Figure A1 in Appendix A). The calculated bond length (dC=O = 1.172 Å for CO2) and bond angle (θOCO = 179.96° for CO2) are in agreement with the findings of previous studies [45,46]. The initial phase of the catalytic conversion of CO2 involves the adsorption mechanism, where the molecule can either undergo physisorption or chemisorption onto the catalyst [26,45,47]. In the chemisorbed state, CO2 displays elongated C-O bonds and a decreased O-C-O bond angle, shifting from a linear to a bent configuration. This transformation implies that CO2 activation occurs due to electron transfer from the metal catalyst to the CO2 molecule’s π molecular orbitals [7,48]. Conversely, in the physisorbed state, CO2 retains its gas-phase characteristics, with an O-C-O bond angle of 180° and a C-O bond length of 1.18 Å. To provide a more systematic investigation, we carried out additional calculations on the Ag5 and Cu5 mono-clusters interacting with CO2, as shown in Figure A2. The relative energies of optimized Cu5@CO2 structures (Figure A2b) illustrate that structure 1 is the most stable, with an adsorption energy (0.70 eV) comparable to that of the Cu4Ag1@CO2 cluster. In contrast, the Ag5@CO2 configurations (Figure A2a) display a slightly different stability trend, with Structure 1 also being the most favorable but exhibiting different adsorption characteristics compared to Cu5@CO2. The pDOS and CDD plots for Ag5@CO2 and Cu5@CO2 are presented in Figure A2c,d, respectively. The pDOS for Ag5@CO2 signifies a less pronounced interaction between the Ag d-orbitals and CO2 at lower energy states (~−4.5 eV), while the Cu5@CO2 cluster reveals remarkable hybridization between the Cu d-orbitals and the CO2 orbitals in the range between −1 and −5 eV. This difference in electronic structure agrees with the CDD plots, where Cu5@CO2 exhibits more charge density transfer to the CO2 molecule than Ag5@CO2, similar to the trend observed in Cu4Ag1@CO2. Comparing these results with the Cu4Ag1@CO2 cluster, we observe that Cu5@CO2 possesses a higher charge transfer capacity, which could suggest enhanced CO2 activation. However, the mixed composition in Cu4Ag1@CO2 favors both stability and effective electron transfer, giving an ideal balance for CO2 activation. This systematic comparison emphasizes the role of Cu and Ag content in impacting the adsorption behavior and electronic attribution, aiding to enhance the design principles for efficient CO2 reduction catalysts.
Considering the most stable configuration of the Cu4Ag1 bimetallic cluster, i.e., structure 1 in Figure 2a, we placed the CO2 molecule at different sites of the cluster as displayed in Figure 6a, and we found that structure 1 is the most stable configuration compared to the metastable state, specifically, structure 2, with an energy difference of roughly 0.5 eV. In structure 1, the CO2 is highly adsorbed on the bimetallic cluster with an adsorption energy of 0.95 eV, implying a chemisorption process. Our result showed that the systems responsible for CO2 activation consistently illustrate effective adsorption and significant charge transfer of around 0.6 e¯ from the Cu4Ag1 to the CO2 molecule. This process leads to the transformation of the CO2 molecule from a linear to a bent structure (θOCO = 135.5° for CO2), along with an elongation of the C-O bonds (dO-C-O = 1.29, 1.24 Å), which is in good agreement with a previously published study on bimetallic CuNi nanoparticles [45]. The CO2 molecule favorably binds to the top sites of the Cu4Ag1 bimetallic cluster, forming bonds with two Cu atoms (dO-Cu = 1.95 Å and dC-Cu = 1.96 Å). The adsorption of CO2 on Cu4Ag1 bimetallic clusters results in modifications in the geometric structures of the clusters, with an observed Cu−Cu average bond length of approximately 2.41 Å. The outcomes indicate that bonding the CO2 to two Cu atoms in Cu4Ag1 bimetallic clusters can enhance their surface activity.
Figure 6b shows the adsorption of CO2 on bimetallic Cu1Ag4 clusters. In this case, the CO2 molecule exhibits a preference for the “top” site, yet no bonds form between CO2 and silver atoms. Considering the most stable configuration of the Cu1Ag4 bimetallic cluster as depicted in structure 1 in Figure 2b, we positioned the CO2 molecule at various sites on the cluster, as demonstrated in Figure 6b. Our estimates revealed that structure 1 remains the most stable structure, in contrast to the metastable state, structure 2, with an energy difference of nearly 0.13 eV. In structure 1, although the CO2 molecule exhibits high adsorption on the bimetallic cluster with an adsorption energy of 0.87 eV, the bond angles (θOCO = 179.67° for CO2) and bond lengths of CO2 (dO=C=O = 1.18, 1.19 Å) closely resemble those found in isolated CO2. This could be explained by slight charge transfer to the CO2 molecule, which is nearly 0.03 e¯. The calculated average bond length of Ag-Ag was observed to be approximately 2.63 Å, which is in good agreement with the study of Ag5@CO2 [49]. It can be concluded that although the adsorption energies of CO2 are comparable for both systems, the degree of electron transfer to CO2 varies substantially, which distinguishes the process of CO2 activation. Further comparison of both systems’ geometric structural parameters is illustrated in Table 1.
To elucidate the factors affecting the chemisorption of CO2 on the Cu4Ag1 bimetallic cluster, we performed calculations to determine the pDOS as demonstrated in Figure 7a. As illustrated in Figure 7a and supported by reference [50], Cu4Ag1@CO2 exhibits delocalized orbitals around the Fermi level, which is clearly evident by the hybridization of sp orbitals of Cu and O. This overlap can facilitate substantial electron transfer from these clusters to CO2, which in turn stabilizes the chemisorbed state. In contrast, the overlap orbitals of Cu1Ag4 with O (p) orbitals of the CO2 molecule are characterized by considerably lower energy levels from the Fermi level, which can be observed at −4.7 eV in Figure 7b. Furthermore, our investigation into the electron excitation of Cu4Ag1 clusters shows that the energy level of the lowest unoccupied molecular orbital (LUMO) in CO2 closely matches the excitation energies of p electrons in Cu4Ag1 clusters (see states at 3.2 eV in Figure 7a). This alignment enables the transfer of excited hot electrons from the LUMO of Cu4Ag1 clusters to the LUMO of CO2 owing to the strong coupling between their orbitals. The resultant electron-transfer state possesses a chemisorption state, which drastically reduces the energy barriers for breaking the C-O bond and forming CO [51]. While in the Cu1Ag4 system, we observed no hybridized LUMO states of the CO2 molecule with the Cu1Ag4 cluster, implying a stable state of the CO2 molecule. In addition, the excess charge observed near the O atom signifies its predominant role in the adsorption of the CO2 molecule rather than the C atom [46]. Charge density maps, as shown in Figure 7a, corroborate our results regarding the electron transfer from the clusters to the CO2 molecule.
2.4. Adsorption of CO2 on Cu3Ag2 and Cu2Ag3 Clusters
Upon analyzing the most stable configuration of the Cu3Ag2 bimetallic cluster, structure 1 in Figure 4a, we explored the positioning of the CO2 molecule at various sites on the cluster, as shown in Figure 8a. Our calculation indicates that structure 1 represents the most stable cluster, with an energy difference of about 0.40 eV compared to the less stable structure 5. Structure 1 shows strong CO2 adsorption on the bimetallic cluster, with an adsorption energy of 0.81 eV, indicative of a chemisorption mechanism. Our results further demonstrate that systems involved in CO2 activation typically reveal strong adsorption and moderate electron transfer, approximately 0.14 e−, from the Cu3Ag2 bimetallic cluster to the CO2 molecule. This interaction induces a transformation in the CO2 molecule from a linear to a bent geometry (θOCO = 137.65° for CO2) and elongates the C-O bonds (dO=C=O = 1.27, 1.26 Å). The CO2 molecule selectively binds to the sites on the Cu4Ag1 bimetallic cluster with the highest adsorption, forming chemical bonds with two Cu atoms (dO-Cu = 2.10 Å and dC-Cu = 2.07 Å). The adsorption of CO2 on the Cu3Ag2 cluster results in geometric alterations, with an average bond length of approximately 2.44 Å between Cu atoms. These results imply that the surface activity of Cu3Ag2 bimetallic clusters can be enhanced by the bonding of CO2 to two Cu atoms.
Figure 8b shows the adsorption process of CO2 on bimetallic Cu2Ag3 clusters. In this scenario, after examining the most stable arrangement of the Cu2Ag3 bimetallic cluster, which is referred to as structure 1 in Figure 4b, we considered the placement of the CO2 molecule at several locations on the cluster, as shown in Figure 8b. The result indicates that structure 1 is the most stable configuration, showing an energy difference of roughly 0.13 eV when compared to the less stable structure 4. Structure 1 exhibits a high adsorption for CO2, as evidenced by its strong adsorption with an energy of 0.81 eV, implying a chemisorption process. Our result also reveals that systems responsible for CO2 activation generally demonstrate substantial adsorption and a high level of electron transfer, around 0.53 e−, from the Cu2Ag3 bimetallic cluster to the CO2 molecule. This interaction causes the CO2 molecule to transition from a straight shape to a curved one (θOCO = 140.36° for CO2) and lengthens the C-O bonds (dO=C=O = 1.24, 1.26 Å). The CO2 molecule exhibits a strong preference for binding to the top site on the Cu2Ag3 bimetallic cluster, making chemical bonds with one Cu atom and one Ag atom (with bond distances of dO-Cu = 2.10 Å, dO-Ag = 2.45 Å, and dC-Cu = 2.03 Å). The results indicate that the surface reactivity of Cu2Ag3 bimetallic clusters can be improved through the bonding of CO2 to the top site of the cluster. Table 2 presents a detailed comparison of the geometric structural parameters of both systems.
To elucidate the factors influencing the chemisorption of CO2 on the Cu3Ag2 bimetallic cluster, we performed calculations to determine pDOS, as shown in Figure 9a. Figure 9a indicates that Cu3Ag2@CO2 exhibits delocalized states near the Fermi level, evidenced by the hybridization of Ag (s), Cu (d), and O (p), and C (p) orbitals. This overlapping orbital facilitates charge transfer from the cluster to CO2, thereby stabilizing the chemisorbed state. Similarly, the orbitals of the Cu2Ag3 cluster overlap with the CO2 orbitals, namely Cu (p), Cu (d), O (p), and C (p), as observed at Fermi energy in Figure 9b. This hybridization allows charge transfer from the cluster to CO2. Additionally, our analysis of electron excitation in Cu3Ag2 clusters shows that the energy level of the LUMO in CO2 closely matches the excitation energies of p electrons in Cu3Ag2 clusters (see states at 3.2 eV in Figure 9a). The alignment between the LUMO of Cu4Ag1 clusters and the LUMO of CO2 facilitates effective transmission of excited hot electrons due to strong orbital coupling. This charge-transfer state adopts a chemisorption structure, effectively lowering the energy barriers for C-O bond dissociation and CO formation. Similarly, in the Cu1Ag4 system, the energy level of the LUMO in CO2 aligns closely with the excitation energies of the p electrons of Cu and s electrons of Ag, particularly at the states around 3.5 eV, as illustrated in Figure 9b. This alignment between the LUMO of Cu4Ag1 clusters and the LUMO of CO2 enables the efficient transfer of excited hot electrons, supported by a strong orbital interaction. Furthermore, the presence of excess charge near the O atom suggests its predominant role in CO2 adsorption, as opposed to the C atom. The CDD maps presented in Figure 9a,b corroborate our results regarding electron transfer from the clusters to the CO2 molecule.
Source link
Turki Alotaibi www.mdpi.com