† Corresponding author. E-mail:
In this study, truncated octahedron (TO) structure is selected for further analysis and we focus on 38-atom Pd–Pt–Ag trimetallic nanoalloys. The best chemical ordering structures of PdnAg32 – nPt6 trimetallic nanoalloys are obtained at Gupta level. The structures with the lowest energy at Gupta level are then re-optimized by density functional theory (DFT) relaxations and DFT results confirm the Gupta level calculations with small shifts on bond lengths indicating TO structure is favorable for 38-atom of PdnAg32 – nPt6 trimetallic nanoalloys. The DFT excess energy analysis shows that Pd8Ag24Pt6 composition has the lowest excess energy value in common with excess energy analysis at Gupta level. In Pd8Ag24Pt6 composition, eight Pd atoms are central sites of 8 (111) hexagonal facets of TO, 24 Ag atoms locate on surface, and 6 Pt atoms locate at the core of the structure. It is also obtained that all of the compositions except Pd18Ag14Pt6 and Pd20Ag12Pt6 exhibit a octahedral Pt core. Besides, it is observed that there is a clear tendency for Ag atoms to segregate to the surface and also Pt atoms prefer to locate at core due to order parameter (R) variations.
Recently, the focus of clusters has been shifting toward nanoalloys consisting of two or more different metals within nanometer size.[1] The size-dependent tunability of nanoalloys offers an excellent opportunity for tailoring new nanoscale materials.[2] Besides, nanoalloys offer the opportunity for tailoring the structures and properties of materials through the choice of atom type and composition.[3,4] Alloying of two or more metals at nanosize is of interest especially for catalysts since the combination of metals can increase the activity and/or the selectivity of catalysts.[5] Since the activity and selectivity are closely associated with structure, their surface structures, compositions, and segregation properties are of particular interest.[6] One promising route to determine and increase the catalytic performance is properly controlling of the size, structure, and composition of nanoalloys.[7] In addition to the geometry and size of the particles, catalytic performance of nanoalloys depends on the chemical ordering.[8] For this reason, a theoretical study of nanoalloys is important in order to find and develop the properties which they attain when their geometric structure and chemical composition varies.[2]
The reason for concentrating here on truncated octahedron (TO) structure is that the selected motif possesses high symmetry,[9] and frequently adopted in theoretical and experimental studies.[10] The 38-atom TO was chosen as model for the trimetallic nanoalloys due to the high symmetry of the parent TO structure (Oh)[11] with 8 (111) hexagonal and 6 (100) square faces.[12] Besides, the size is interesting since it corresponds to magic number of atoms for the truncated octahedron[13] and it is the only size below 50 atoms at which fcc structures are in competition with icosahedral fragments and amorphous structures.[14] Furthermore, it was found that TO structures typically dominate 38-atom nanoclusters in several studies of literature and also they were obtained as the global minimum (GM) for many clusters for the Gupta many body potential,[4,15,16] as well as the other many-body and pair potentials.[17–19]
There have been many theoretical studies of mono and bimetallic nanoclusters with 38-atom TO structure to investigate the relationship between geometric structure, chemical ordering and segregation tendency.[2,4–11,14,15,20–25] Zhu et al.[7] investigated Pd–Au nanoalloys for 38-atom TO. Demiroğlu et al.[11] performed a density functional theory (DFT) study to investigate the effect of the TiO2 support on the structures of 38-atom AuRh nanoalloys and in another study nanoscale mixing properties of 38-atom Au–Rh were investigated.[8] Roy et al.[20] studied 38-atom Pd–Pt, Ag–Au, Pd–Au, and Ag–Pt binary clusters combining empirical potential (EP) and DFT calculations. Rodrigues et al.[2] calculated the global optimization of CunAum (n + m = 38) clusters. A study of chemical ordering in 38-atom Pd–Ir nanoalloys were reported by Davis et al.[5] Cerbelaud et al.[21] determined the optimal chemical ordering for binary AgAu cluster with 38-atom. Curley et al.[6] studied the Ag–Au bimetallic nanoalloy clusters with 38 atoms using Gupta many-body potential. For chemical ordering in AgAu TO nanoalloys of 38 atoms, Rapetti et al.[14] also reported a DFT global optimization study. Darby et al.[4] reported that Cu38 and Au38 clusters have a TO structure. Molayem et al.[22] found global minimum structures based on TO for some Cu–Ag nanoalloys with 38 atoms. A theoretical investigation of 38-atom CuPt clusters was performed by Guerrero-Jordan et al.[23] using a genetic algorithm. Pittaway et al.[24] also used genetic algorithm for global optimization of 38-atom PdAu clusters with TO structure. Fan et al.[10] investigated the structural stability and chemical ordering of 38-atom TO Pd–Ir nanoalloys. Melting mechanism of Pd24Pt14 nanoalloy with TO was determined by Oderji et al.[25] Negreiros et al.[9] analyzed the kinetics of chemical ordering in a Ag–Pt TO nanoalloy particle via first principles simulations. TO structure and energetics of Ni, Ag, and Au nanoclusters of size 38 atoms were investigated by Michaelian et al.[15]
Despite of these investigations on the monometallic and bimetallic nanoclusters with 38-atom TO structure, there are not sufficient data about trimetallic 38-atom nanoclusters. Pacheco-Contreras et al.[13] investigated Ag–Au–Pt trimetallic clusters contain a total of 38 atoms at Gupta and DFT levels. Wu et al.[26] investigated geometrical and energetic properties in 38-atom Au–Pd–Pt clusters. The structrures of Cu8AunPt30 – n clusters were analyzed by Wu et al.[27] Besides, many above-mentioned investigations performed simulation searches by combining EP and DFT approaches. Since there is still lacking a thorough study presenting the relationship between geometric structure, chemical ordering, and segregation tendency of trimetallic nanoalloys with 38-atom TO, we have focused on 38-atom Pd–Pt–Ag trimetallic nanoalloys with TO structure in this study. Although some investigations[28,29] on clusters present comparisons with available experimental results, it is not always possible to directly compare theoretical results with experimental data for clusters. To study small clusters with rather accurate methods from the theoretical point of view, DFT calculations could be an alternative choice. Thus, we have performed DFT re-optimization for the best chemical ordering structures of PdnAg32 – nPt6 trimetallic nanoalloys obtained at Gupta level.
In order to model the metallic bonding between Pt, Pd, and Ag atoms in Pd–Ag–Pt trimetallic naoalloys, Gupta potential[30,31] has been used. Gupta potential can be written in terms of repulsive (
The best chemical ordering structures of PdnAg(32 – n)Pt6 nanoalloys with TO structure were carried out using Monte Carlo Basin–Hopping algorithm.[34,35] The best chemical ordering structures obtained at Gupta level were chosen as initial configurations and re-optimized DFT searches were performed for the chemical ordering in PdnAg(32 – n)Pt6 nanoalloys. The DFT calculations with Perdew–Burke–Ernzerhof (PBE) exchange–correlation functional[36] were made by using the Quantum ESPRESSO package.[37,38] In Quantum ESPRESSO, the energy cutoff for wavefunctions was used as 45.0 Ry (1 Ry = 13.6056923(12) eV) and the cutoff energy for charge density was used as 448.0 Ry.
In the current study, chemical ordering in 38-atom trimetallic Pd–Ag–Pt nanoalloys with TO structure has been studied and a systematic search was performed for the best chemical ordering structures. Besides, structure and chemical ordering effect on stability of trimetallic nanoalloys are explored. Structural details of trimetallic nanoalloys are analyzed as a function of their chemical compositions. We mainly focus here on the PdnAg(32 – n)Pt6 nanoalloys with fixed six Pt atoms where the fixed number 6 represents inner atoms of the 38-atom TO structure. The high-symmetry structure is taken as initial atomic configurations.
The best chemical ordering structures obtained at Gupta level for PdnAg(32 – n)Pt6 nanoalloys are presented in Fig.
In the nanoscale, the energetic stability among a family of nanoalloys is determined by calculating the excess energy (ΔEexc). For trimetallic PdnAg(32 – n)Pt6 nanoalloys with fixed size but different compositions excess energy is defined as[39]
In order to provide more useful structural information about the chosen nanoalloy system, Gupta and DFT approaches were combined. The best chemical ordering structures at Gupta level were recalculated with DFT relaxations to crosscheck the lowest energy structures. Since there is not occurrence of any discrepancies on nanoalloy structures, it can be interpreted as DFT results confirm the Gupta level calculations with small shifts on bond lengths. Also, DFT calculations indicate that the truncated octahedron structure is indeed the lowest energy ones for PdnAg(32 – n)Pt6 nanoalloys. Figure
The structure of the most stable composition Pd8Ag24Pt6 for PdnAg32 – nPt6 nanoalloys is illustrated in detail in Fig.
In order to better discuss about the atomic mixing degree of different type atoms in Pd–Ag–Pt nanoalloys, the order parameter (RA)[40] is adopted. RA can be defined by the average distance of a type of atoms from the center of a cluster, i.e.,
Figure
The strain effects are more important in the bimetallic and trimetallic systems, because of the size mismatch between different atomic species. Also, the strain becomes different in different geometries, indicating that the chosen geometry and chemical ordering are interrelated. Since nanoparticles with strained surfaces may present important catalytic properties, we decided to investigate the local atomic pressure values for Pd–Ag–Pt nanoalloys. The local pressure Pi acting on atom i is proportional to the trace of the tensor given as follows:[41,42]
Figure
Figure
In this study, we have performed a systematic investigation of PdnAg32 – nPt6 trimetallic nanoalloys and we have focused on 38-atom Pd–Pt–Ag trimetallic nanoalloys with truncated octahedron (TO) structure. The best chemical ordering structures were obtained using Monte Carlo basin-hopping algorithm within Gupta potential. The structures with the lowest energy at Gupta level are then re-optimized by DFT relaxations and DFT results confirm the Gupta level calculations with small shifts on bond lengths, indicating TO structure is favorable for 38-atom of PdnAg32 – nPt6 trimetallic nanoalloys.
In PdnAg(32 – n)Pt6 trimetallic nanoalloys, all the compositions except Pd18Ag14Pt6 and Pd20Ag12Pt6 nanoalloys exhibit a octahedral Pt core. From the order parameter variations of trimetallic PdnAg32 – nPt6 nanoalloys, it was observed that Ag atoms exhibit a clear tendency to segregate to the surface. DFT level investigations showed that two chemical orderings corresponding Gupta level minima with 5 Pt atoms at core (for Pd18Ag14Pt6 and Pd20Ag12Pt6) are energetically favorable than the chemical orderings with 6 Pt atoms at core. Moreover, DFT and Gupta excess energy analysis show that Pd8Ag24Pt6 composition has the lowest excess energy value which corresponds to more stable structure. In Pd8Ag24Pt6 composition, 8 Pd atoms are central sites of 8 (111) hexagonal facets of TO, 24 Ag atoms locate on surface and 6 Pt atoms locate at the core of the structure. Furthermore, it was observed that 8 Pd atoms at central sites of hexagonal facets exhibit more positive pressure than the atoms at central sites of bimetallic Ag32Pt6 and Pd32Pt6 nanoalloys in relation to mismatch between three different atomic species and also chemical ordering.
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] | |
[24] | |
[25] | |
[26] | |
[27] | |
[28] | |
[29] | |
[30] | |
[31] | |
[32] | |
[33] | |
[34] | |
[35] | |
[36] | |
[37] | |
[38] | |
[39] | |
[40] | |
[41] | |
[42] |