Lorena
Vega
a,
Hristiyan A.
Aleksandrov
*b,
Riccardo
Farris
a,
Albert
Bruix
a,
Francesc
Viñes
a and
Konstantin M.
Neyman
*ac
aDepartament de Ciència de Materials i Química Física & Institut de Química Teòrica i Computacional (IQTCUB), Universitat de Barcelona, 08028 Barcelona, Spain
bFaculty of Chemistry and Pharmacy, University of Sofia, 1126 Sofia, Bulgaria. E-mail: haa@chem.uni-sofia.bg
cICREA (Institució Catalana de Recerca i Estudis Avançats), 08010 Barcelona, Spain. E-mail: konstantin.neyman@icrea.cat
First published on 6th August 2021
Bimetallic alloys are actively investigated as promising new materials for catalytic and other energy-related applications. However, the stable arrangements of the two metals in prevailing nanostructured systems, which define their structure and surface reactivity, are seldom addressed. The equilibrium chemical orderings of bimetallic nanoparticles are usually different from those in the corresponding bulk phases and hard to control experimentally, which hampers assessment of the relations between composition, structure, and reactivity. Herewith, we study mixtures of platinum—an essential metal in catalysis—alloyed with coinage metals gold, silver, and copper. These systems are interesting, for instance, for reducing the costly Pt content and designing improved multifunctional catalysts, but the chemical orderings in such mixtures at the nanoscale are still debated. We therefore explore chemical orderings and properties of Pt-containing nanoalloys by means of a topological method based on density functional calculations. We determine the lowest-energy chemical orderings in 1.4 to 4.4 nm large Pt–Au, Pt–Ag and Pt–Cu particles with different contents of metals. Chemical ordering, bonding, and charge distribution in the nanoparticles are analyzed, identifying how peculiar structural motifs relevant for catalysis and sensing applications, such as monometallic skins and surface single-atom sites, emerge. We compare these results with previous data for the corresponding Pd-based particles, identifying trends in chemical ordering, deepening understanding of the behaviour of catalytically relevant bimetallic compositions, and establishing appropriate models for studying the bimetallic nanoalloys.
The size, shape, and composition of bimetallic NPs can be quite well controlled by the preparation conditions. At variance, such elusive degree of complexity as chemical (or atomic) ordering, i.e. a distribution pattern of metal atoms of two types among lattice positions of an alloy NP, is hard to precisely control experimentally. The reactivity of metal NPs is directly related to their surface arrangement. The chemical ordering defines what types and how many surface sites are exposed by a bimetallic NP of a given size, shape, and composition. This is key information for catalysis, sensing, and many other applications. Yet, it remains barely accessible experimentally at the atom-by-atom level even for the most modern structure characterization techniques.
Computational modeling using Density Functional Theory (DFT) can provide detailed information about the structure and properties of bimetallic NPs, complementing experimental data. DFT calculations of bimetallic NPs with over a hundred atoms and sizes of ∼1.5 nm, which are required to realistically represent larger particles dealt with in catalysis,3–5 are feasible since two decades.6 However, the presence of more than one type of atoms in nanoalloys severely hinders their comprehensive DFT simulation, often restricting it to quite small particles and considering only several chemical orderings (homotops).7–12 In fact, a direct search for the equilibrium chemical ordering in a ∼2 nm large bimetallic crystallite comprising ∼200 atoms requires calculating energies of a colossal number of 1050 homotops (including symmetry-equivalent ones),13 which is excessive for any computational method.
This challenge can be dealt with by a Topological (TOP) approach,14,15 which enables determining equilibrium chemical orderings in bimetallic nanocrystallites containing 102–104 atoms of different metals across the Periodic Table from a small number of DFT calculations. Briefly, the TOP method divides all homotops of a bimetallic nanocrystallite with a given stoichiometry and shape into groups with the same topologies. Definition of the latter depends on how atomically detailed the resulting ordering needs to be. For instance, for studying the catalytic activity of a bimetallic NP comprising m + n atoms it is essential to know how many active surface sites of each type the NP exposes. These data are related to the propensity of M′ and M atoms to segregate on the surface and define which atoms occupy surface positions of the NP with different coordination numbers, e.g. in corner, edge, and terrace sites. Since all homotops of the NP under scrutiny share the same crystal lattice and composition, it suffices to specify atomic positions for just one of the two metals. The types of the exposed surface sites and their abundance are also affected by the propensity of the metals M′ and M to mix, which is reflected by the number of heterometallic bonds (nearest-neighbor atom pairs) M′–M formed. For surface reactivity studies it is often sufficient to specify ordering patterns in bimetallic nanoalloys by the topology defined solely by the number of M′–M bonds and surface M atoms in the corner/vertex (NMCORNER), edge (NMEDGE), and facet/terrace (NMFACET) positions. Hereafter, topologies of NPs are designated as ; to specify a particular homotop of the considered topology, its EDFT value in eV is added to the above designation, resulting in the designation . Energies of all homotops with a given topology are represented in the TOP method by the energy of just one of the homotops, chosen arbitrarily.14
Energy difference, ΔETOPij, between any two homotops i and j of a bimetallic NP reads:14,16
(1) |
Here, the energies ε are contributions of either a M′–M bond or an M atom located in a certain outer position of the NP to its topological energy ETOP. The terms ε quantify either the surface segregation energy of metal atoms M′ and M or the energy gained or lost upon metals mixing to form a pair M′–M of the nearest-neighbor atoms. These terms are calculated by fitting eqn (1) to DFT energies EDFT of a series of homotops of a bimetallic NP with diverse topologies. Using thus obtained TOP expressions a comprehensive ordering screening is performed by means of Monte Carlo (MC) simulations to find the topology—a set of the N values—of the most energetically stable homotops of the NP along with topologies of less stable homotops. Then, atomic positions in the selected low-energy homotops are optimized by DFT. This allows examining accuracy of the energies ε and eqn (1) overall. The resulting latter expressions can be used to simulate the chemical orderings in bimetallic NPs containing thousands atoms, not accessible for DFT calculations.
Successfully determined chemical orderings in NPs of various metal combinations, such as Pd–Au,13–17 Pd–Ag,14 Pd–Cu,14 Pd–Zn,14 Pd–Rh,16 Pt–Ag,18 Pt–Co,19,20 Pt–Ni,21 Pt–Sn,22 and Ni–Cu,23 revealed efficiency and broad applicability of the TOP method. This study deals with bimetallic nanoalloys of essential for catalysis metal Pt with quite inert coinage metals Au, Ag, and Cu. These combinations of metals are actively studied in the rapidly developing field of single-atom catalysis24,25 as single-atom alloy catalysts,26 where individual Pt atoms surrounded by coinage atoms can act as very selective catalytic centers.
Among the aims of the present study are:(i) quantifying by DFT calculations the chemical orderings and surface segregation in up to ∼2 nm large Pt–Au, Pt–Ag and Pt–Cu crystallites at different concentrations of the metals; (ii) analyzing the structure, bonding, and atomic charges in these nanoparticles compared to the analogous Pd-containing nanoalloys studied earlier; (iii) deepening the knowledge of the accuracy and applicability of the Topological approach; (iv) identifying equilibrium chemical orderings in Pt–Au, Pt–Ag and Pt–Cu particles at sizes over 4 nm commonly found in catalysts, also at elevated temperatures.
In line with our DFT calculations of other nanoalloy particles,13,14,16,21 here the Pt–X (X = Au, Ag, Cu) NPs were also represented by truncated octahedrons with the fcc lattice. Ca. 1.4 nm large 140-atomic Pt70Au70, Pt70Ag70, and Pt70Cu70 NPs, see Fig. 1, were considered for comparison with the results for Pd-based Pd70Au70, Pd70Ag70, and Pd70Cu70 analogs.14 1.7–1.8 nm large 201-atomic NPs Pt–Au, Pt–Ag and Pt–Cu with Pt:X 1:3, 1:1, and 3:1 compositions, see Fig. 2, were modelled to study the size, shape, and composition effects. Placing these 140- and 201-atom NPs in periodically repeated cubic cells of 2.3 × 2.3 × 2.3 and 2.5 × 2.5 × 2.5 nm3 allowed separation >0.7 nm between them. At such distances interaction of metal particles with their periodic images is negligible for the purpose of this work.35,36
Fig. 2 Chemical orderings of 1.7–1.8 nm large 201-atomic truncated-octahedral fcc nanoparticles of the compositions Pt:X (X = Au, Ag, Cu):1:3 – Pt51X150, 1:1 – Pt101X100, 3:1 – Pt151X50 with the lowest DFT energies. Particles images are split in two parts for better visualization. Colors of atoms as in Fig. 1. |
Nanoparticle | ε Pt-XBOND | ε XCORNER | ε XEDGE | ε XFACET | δ | ΔE | N FIT |
---|---|---|---|---|---|---|---|
a 95% confidence interval of ε given as, e.g. 19+4−2 denotes a range from 17 meV to 23 meV. b When several homotops were optimized by DFT for one of the selected ≥10 low-energy topologies, see Section 3.6, all these EDFT values were also used in the calculations of δ. | |||||||
Pt70Au70 | 19+4−2 | −619+62−48 | −377+44−44 | −256+88−69 | 470 | 0 | 35 |
Pd70Au70 | −13+4−6 | −404+76−72 | −301+52−77 | −200+52−64 | 115 | 26 | 32 |
Pt70Ag70 | 11+10−13 | −625+122−185 | −336+72−185 | −195+75−58 | 358 | 0 | 45 |
Pd70Ag70 | −1+2−2 | −361+50−68 | −289+78−129 | −163+43−64 | 150 | 29 | 53 |
Pt70Cu70 | −35+4−4 | −27+47−44 | 182+56−54 | 344+43−36 | 879 | 415 | 65 |
Pd70Cu70 | −26+5−5 | 95+36−33 | 147+46−45 | 183+42−40 | 360 | 171 | 127 |
Pt151Au50 | 21+1−1 | −507+51−62 | −543+16−18 | −431+30−43 | 114 | 354 | 101 |
Pt101Au100 | 21+7−6 | −530+136−107 | −492+59−56 | −335+83−95 | 456 | 0 | 68 |
Pt51Au150 | 15+14−15 | −558+49−34 | −547+53−84 | −259+47−92 | 279 | 198 | 44 |
Pt151Ag50 | 32+11−12 | −396+108−135 | −380+95−90 | −237+103−95 | 461 | 176 | 90 |
Pt101Ag100 | 16+9−10 | −499+95−112 | −466+87−83 | −308+77−91 | 493 | 65 | 68 |
Pt51Ag150 | 7+5−5 | −408+118−158 | −511+79−68 | −240+49−65 | 169 | 43 | 99 |
Pt151Cu50 | −25+9−13 | 267+30−22 | 342+14−14 | 372+42−44 | 784 | 204 | 165 |
Pt101Cu100 | −43+6−5 | 15+60−72 | 208+55−68 | 325+34−41 | 576 | 284 | 87 |
Pt51Cu150 | −54+14−15 | 134+73−78 | 184+151−93 | 259+65−59 | 239 | 0 | 40 |
Particle | N Pt-XBOND | N XCORNER | N XEDGE | N XFACET | N X100 | E DFTDFTmin (eV) | E DFTTOPmin (eV) |
---|---|---|---|---|---|---|---|
a The overall topological numbers NBOND, NCORNER, NEDGE, NFACET, and NINSIDE are 636, 24, 24, 48, 44 (Pt70X70), and 948, 24, 36, 62, 79 (Pt201−nXn), respectively. | |||||||
Pt70Au70 | 196/196 | 24/24 | 24/24 | 22/22 | −588.108 | −588.108 | |
Pt70Ag70 | 196/196 | 24/24 | 24/24 | 22/22 | −561.294 | −561.294 | |
Pt70Cu70 | 308/340 | 23/24 | 8/12 | 0/0 | −626.920 | −626.610 | |
Pt151Au50 | 215/214 | 23/20 | 25/30 | 2/0 | 2/0 | −978.389 | −978.283 |
Pt151Ag50 | 222/215 | 24/21 | 26/29 | 0/0 | 0/0 | −962.454 | −961.772 |
Pt151Cu50 | 404/412 | 0/0 | 0/0 | 0/0 | 0/0 | −1015.683 | −1015.214 |
Pt101Au100 | 267/267 | 24/24 | 36/36 | 40/40 | 5/5 | −859.505 | −859.505 |
Pt101Ag100 | 274/270 | 24/24 | 36/36 | 40/40 | 6/4 | −822.032 | −821.660 |
Pt101Cu100 | 525/560 | 24/24 | 10/8 | 3/8 | 0/0 | −914.377 | −913.767 |
Pt51Au150 | 224/210 | 24/24 | 36/36 | 62/62 | 6/6 | −730.046 | −729.869 |
Pt51Ag150 | 220/214 | 24/24 | 36/36 | 62/62 | 6/6 | −669.026 | −668.827 |
Pt51Cu150 | 418/418 | 24/24 | 36/36 | 18/18 | 0/0 | −799.627 | −799.627 |
The following literature data are relevant for the present study of the Pt–X nanoalloys:(i) the surface energies calculated for close-packed surfaces of the pure metals are 1.03 (Pt), 0.72 (Au), 0.58 (Ag), and 0.77 eV per atom (Cu);37 (ii) calculated solutes/hosts surface segregation energies of monoatomic impurities (solutes) in fcc(111) metal hosts are −0.36 (Au/Pt), 0.34 (Pt/Au), −0.27 (Ag/Pt), 0.34 (Pt/Ag), 0.32 (Cu/Pt), and −0.04 eV (Pt/Cu), where the negative sign indicates stabilizing surface segregation of the solute;37 (iii) Pt atoms in the bulk alloys are immiscible with Au38–41 and Ag40 atoms, but miscible with Cu40,42–44 atoms; and (iv) relative sizes of metal atoms,45 139 (Pt), 144 (Au), 144 (Ag), and 128 pm (Cu), affect the propensity of smaller atoms to surface segregate in compressively strained bimetallic particles40,46–48 so that the surface segregation of Cu is suppressed in Pt–Cu nanoalloys despite Cu has a lower surface energy than Pt (see above).
The usefulness of the TOP approach and the meaning of terms in eqn (1) can be illustrated by comparing the TOP and DFT segregation energies. A homotop of Pt51Au150 with almost lowest TOP energy, Pt51Au150_212.24.36.62–729.864 (see ESI†), is characterized by NPt-AuBOND = 212, NAuCORNER = 24, NAuEDGE = 36 and NAuFACET = 62 and EDFT = −729.864 eV (homotops are labeled in this way throughout the present article). Exchange of an atom Au from center of a {111} facet with a near-neighbor subsurface atom Pt (each of the exchanged atoms Au number 39 and Pt number 198 formed 3 Pt–Au bonds) results in the homotop Pt51Au150_223.24.36.61–729.507 (see ESI†), where, the surface Pt atom forms 7 Pt–Au bonds and the Au atom moved to the subsurface forms 10 Pt–Au bonds. Its energy increase with respect to the pristine homotop is , using eqn (1) and energies ε in Table 1. The DFT energy increase 357 meV is close to the ΔETOP value in terms of the precision δ(EDFT − ETOP) = 279 meV (Table 1). δ(EDFT − ETOP) is twice the residual standard deviation between EDFT and ETOP energies for homotops of ≥10 test low-energy topologies, according to which the respective ΔETOP and ΔEDFT values differ by less than δ with the probability ≥95%.14 Exchange of an edge atom Au with a nearby subsurface atom Pt in the pristine homotop models even less favorable surface segregation of inner atoms Pt to edge sites. NPt-AuBOND after this exchange of atom Au number 132 with atom Pt number 176 (see ESI†) increases in the resulting homotop Pt51Au150_219.24.35.62–729.345 for the atom Au from 1 to 6 and for the atom Pt from 5 to 7. The energy increase vs. the pristine homotop, is indeed larger than for the displacement of the inner atom Pt to a {111} facet and reasonably well agrees with ΔEDFT = 519 meV.
These two examples demonstrate that the TOP method, beyond providing adequate relative energies of different chemical orderings of a bimetallic NP, also allows rationalizing the energy differences in such important terms as propensities of their two types of metal atoms to form heterometallic bonds and to occupy differently coordinated surface sites.
Particle | Pt | X | ||||||||
---|---|---|---|---|---|---|---|---|---|---|
6 | 7 | 8 | 9 | 12 | 6 | 7 | 8 | 9 | 12 | |
Pt70Au70 | — | — | — | −0.02 | 0.05 | −0.05 | −0.02 | 0.02 | — | |
Pt151Au50 | −0.09 | −0.07 | −0.08 | −0.02 | 0.02 | −0.05 | −0.03 | — | — | — |
Pt101Au100 | — | — | −0.09 | −0.04 | 0.03 | −0.05 | −0.03 | 0.01 | 0.02 | — |
Pt51Au150 | — | — | — | — | 0.01 | −0.06 | −0.03 | −0.01 | 0.01 | 0.06 |
Pt70Ag70 | −0.17 | −0.04 | 0.05 | 0.10 | 0.10 | — | ||||
Pt151Ag50 | −0.13 | −0.13 | −0.24 | −0.11 | 0.02 | 0.12 | 0.15 | — | — | — |
Pt101Ag100 | — | — | −0.30 | −0.16 | −0.05 | 0.05 | 0.08 | 0.13 | 0.09 | — |
Pt51Ag150 | — | — | — | — | −0.11 | −0.02 | 0.01 | 0.04 | 0.04 | 0.12 |
Pt70Cu70 | — | −0.25 | −0.23 | −0.33 | 0.18 | 0.27 | — | 0.30 | ||
Pt151Cu50 | −0.14 | −0.11 | −0.08 | −0.07 | −0.23 | — | — | — | — | 0.37 |
Pt101Cu100 | — | −0.29 | −0.12 | −0.27 | −0.39 | 0.27 | 0.22 | — | 0.32 | 0.31 |
Pt51Cu150 | — | — | −0.57 | −0.46 | −0.60 | 0.11 | 0.15 | — | 0.19 | 0.19 |
Pt201 | −0.09 | −0.06 | −0.03 | −0.01 | 0.07 | |||||
Au201 | −0.07 | −0.04 | −0.02 | −0.01 | 0.04 | |||||
Ag201 | −0.04 | −0.03 | 0.01 | −0.01 | 0.03 | |||||
Cu201 | −0.05 | −0.02 | 0.02 | −0.01 | 0.03 |
Ordering of the lowest-energy homotops of Pt101Ag100 NP is very similar to those of Pt101Au100 NP, see Fig. 2 and Table 2. All Ag atoms are located on the surface occupying all 24 corner and 36 edge sites and 40 of 62 terrace sites. Pt atoms occupy the remaining terrace sites and all 79 inner sites. This ordering exhibits incomplete Ag outer shell and monometallic Pt core. Qualitatively the same values of energies ε guiding the orderings in 201-atom NPs with other Pt:Ag ratios, Pt51Ag150, and Pt151Ag50, as in the Pt101Ag100 NP, result in completing the Ag skin in the former case and depleting the skin (mainly by the facet atoms) in the latter case. Comparing equilibrium chemical orderings of the 201-atomic Pt–Ag NPs with their Pt–Au analogs, see Table 2, one notices very close similarities, especially at low Pt concentrations. Indeed, topologies of most corresponding Pt–Ag and Pt–Au pairs of NPs differ at most by just small amount of heterometallic bonds contributing up to a fraction of an eV to the NP energy. It is worth mentioning that a peculiar layered chemical ordering L11 was experimentally identified inside monolayer Ag skins of some relatively small Pt–Ag NPs.49 Small energetic preference calculated by DFT of such layering in Pt–Ag NPs18 becomes even smaller in the analogous Pd–Ag NPs and disappears in Pt–Au and Pd–Au NPs that feature monolayer Au skins.50 Thus, the elusive phenomenon of Pt and Pd layering inside the coinage metal skins appears to be uncommon.
At the lowest studied Cu content (3:1 Pt:Cu), all 50 Cu atoms of Pt151Cu50 NP are energetically driven to be located inside the monolayer Pt skin and to form 412 bonds with inner and surface Pt atoms, see Table 2 and Fig. 2. At the 1:1 Cu:Pt content, 60 out of the 100 Cu atoms of Pt101Cu100 NP stay in the core, 24 Cu atoms occupy all corner positions and the remaining 16 Cu atoms are distributed over edge (8 atoms) and facet (8 atoms) positions. The number of Pt–Cu bonds increases to 560. At the lower 1:3 Pt:Cu content, all corner and edge positions of Pt51Cu150 NP are occupied by Cu atoms. Cu atoms are also located in 18 of 62 facet positions and form most of the 79-atom core. Interestingly, see Fig. 2, most of surface Pt atoms are directly neighboring other Pt atoms instead of being completely surrounded by Cu atoms, which would enable maximizing the number of Pt–Cu bonds. This implies that low Pt contents are required to form stable single Pt atoms on the surface of Pt–Cu NPs.
The miscibility of metals M and M′ in a bimetallic NP can be also evaluated by its excess energy vs the energies of the corresponding monometallic Mn+m and particles of the same size and structure.7,55 The excess energy (also known as mixing energy) per atom for a PtmXn particle is calculated as:
Eexc(PtmXn) = {E(PtmXn) − [m/(m + n)]E(Ptm+n) − [n/(m + n)]E(Xm+n)}/(m + n), | (2) |
The DFT excess energies of 201-atom Pt–X NPs are plotted in Fig. 3. The energies for Pt–Au and Pt–Ag NPs with Pt and X atoms that are immiscible in the bulk are spread depending on the concentrations of atoms X from −38 to −65 meV and from −58 to −90 meV, respectively. The corresponding energies for Pt–Cu NPs ranging from −127 to −163 meV are notably more negative, as expected for the miscible Pt and Cu atoms. For all these 201-atomic Pt–X NPs the miscibility in terms of the excess energies reaches maximum close to the 1:1 compositions. This mixing indicator remains nearly the same for the 1:1 140-atomic NPs, see Fig. 3: −54, −80, and −165 meV for Pt70Au70, Pt70Ag70, and Pt70Cu70, respectively.
Fig. 3 DFT excess energies Eexc per atom (see eqn (2)) circles – of Pt201−nXn (X = Au, Ag, Cu; n = 50, 100, 150) nanoparticles and squares – of Pt70X70 nanoparticles. Pt–Au – yellow, Pt–Ag – gray, Pt–Cu – orange. |
The main qualitative difference of changing Pt to Pd in the nanoalloys with Au and Ag, see Table 1, is that, at variance with Pt, Pd becomes slightly miscible with these coinage metals, by −13 and −1 meV per Pd–X bond, respectively. On the other hand, the notable preference of Au and Ag atoms to occupy surface positions remains present in the Pd-based NPs, although the corresponding energies ε for them are significantly smaller than for the analogous Pt-based NPs. The stabilizing effect of the location of Au and Ag atoms in surface positions decreases in both Pt- and Pd-based NPs with increasing coordination number of the surface position. The most favorable are 6-coordinated corner sites, then 7-coordinated edge sites, and finally 9-coordinated sites in the {111} facets. Similarities of the energies ε for the corresponding Pt70X70 and Pd70X70 (X = Ag, Au) NPs result in similar orderings of their lowest-energy homotops. In all four NPs X atoms occupy all 48 corner plus edge sites, while the remaining 22 X atoms are located in the {111} facet sites. Pt and Pd atoms occupy the remaining 26 surface sites of the facets and all 44 positions in the interior of the NPs. The main difference in the orderings comes from the number of heterometallic bonds. The equilibrium Pt70X70 structures feature 196 Pt–X bonds, 31% of overall 636 bonds in the 140-atomic NP, to be compared with significantly larger number of heterometallic bonds in the Pd70X70 structures, 234–262 (37–41%). A consequence of the different numbers of the heterometallic bonds is that most of the {111} facets in the Pt70X70 structures are formed by either only X or only Pt atoms, whereas the individual facets in the Pd70X70 structures contain a mixture of X and Pd atoms, increasing the number of Pd–X bonds.
Miscibility of the coinage and platinum-group metal atoms is favorable in both Pt70Cu70 and Pd70Cu70 NPs, see Table 1. Each heterometallic bond stabilizes the former particle by 9 meV more than the latter. The occupation of surface positions by Cu destabilizes these two NPs, with one exception. Namely, the occupation of corner position of Pt70Cu70 NP by atom Cu instead of atom Pt has a slight stabilizing effect, by −27 meV. The destabilization due to the presence of surface Cu atoms increases with increasing their coordination numbers, though the destabilization values (energy differences between the most stable and the least stable surface positions) are quite different, 371 meV for Pt70Cu70 and 88 meV for Pd70Cu70. It is least favorable for Cu atoms to occupy terrace sites of Pt70Cu70 and Pd70Cu70 NPs. Cu atoms are differently distributed on the surface of Pt70Cu70 and Pd70Cu70 homotops with the lowest ETOP energy. In the Pt70Cu70 NP, 36 surface Cu atoms occupy all 24 corner positions and a half of 24 edge positions, leaving all 48 terrace positions to be occupied by Pt atoms, see Table 2. 32 surface Cu atoms of the Pd70Cu70 NP are quite evenly spread over three types of sites: corner – 12, edge – 14, and terrace – 814 as a result of relatively small energy difference of Cu in these sites. To maximize the number of stabilizing heterometallic bonds the equilibrium Pt70Cu70 and Pd70Cu70 structures are of quite common onion-like type,1,2,56,57 with Pt/Pd-rich core, Cu-rich subsurface shell and monatomic skin shared by Pt/Pd and Cu atoms. These chemical orderings allow formation of 340 Pt–Cu and 382 Pd–Cu bonds.
The present topological approach has been recently employed to analyze in detail the formation of single-atom Pd sites in Pd–Au NPs appealing for catalysis as a function of Pd concentration inside the particles.17 It was shown that no surface Pd atoms emerge at low Pd concentrations in the Au skin at equilibrium chemical orderings. Single-atom surface Pd sites become stable in 201-atom Pd–Au particles when Pd content inside the skin reaches ca. 60% (corresponding to 153 Au and 48 Pd atoms) and further increase of Pd content results in more surface Pd, first mostly as single atoms. Such emergence of surface Pd atoms is triggered by stabilizing Pd–Au bonds reflected in negative εPd-AuBOND energies for Pd–Au NPs.13–15,17 According to εPt-XBOND (Table 1), heterometallic bonds of Pt are stabilizing neither in Pt–Au, nor in Pt–Ag NPs. Thus, the location of Pt atoms is not favored on the surface of Pt–Au and Pt–Ag NPs at the equilibrium orderings, unless the number of Au and Ag atoms in these particles is smaller than the 122 required to form complete coinage-metal skins. However, single-atom Pt sites can be easily created on the surface of Pt–Au and Pt–Ag NPs in the course of catalytic reactions already at low Pt concentration due to energetically preferable interactions of Pt centers with reactants and intermediates compared to Au and Ag centers. This adsorbate-induced surface segregation phenomenon is well known for nanoalloy catalysts.13,16,21,23,25,26,51 In the cases of Pd–Cu and Pt–Cu NPs Pd and Pt atoms show a strong preference to be located in the surface skin and to form stabilizing heterometallic bonds, see Tables 1 and 2. Thus, the presence of Pd and Pt on the surface of these nanoalloys as unique single-atom catalytic sites is conceivable only at very low concentration of the platinum-group metals.
To evaluate the correctness of some of these assumptions we randomly generated ≥10 different homotops for several selected topologies of Pt–Au and Pt–Cu NPs and locally optimized those by DFT; see results in Table S1 (ESI†). The DFT energy splits for the homotops belonging to each of the considered low-energy topologies of 201-atomic Pt–Au NPs are very small, up to ca. 1 meV per atom. The splits slightly increase with growing content of Au atoms (preferably located on the surface), ranging from only 25 meV for the topology Pt151Au50_214.20.30.0 to 286 meV for the topology Pt51Au150_210.24.36.62. DFT energy splits for low-lying topologies of Pt–Cu NPs noticeably increased compared to the corresponding Pt–Au NPs. Noteworthy, the Pt–Cu energy splits also increase with growing content of Pt on the surface, from 153 meV for the topology Pt51Cu150_418.24.36.18 to 930 meV for the topology Pt151Cu50_412.0.0.0, the latter with solely Pt atoms on the surface. We note that the prediction error of ETOPvs. EDFT energies for such low-energy orderings range from 1.1 to 6.8 meV per atom (see Table S1, ESI†), although the errors for high-energy structures are significantly larger. For many practical applications, the up to ca. 7 meV per atom uncertainty in total energy of the Pt151Cu50_412.0.0.0 homotops provided by the present computational setup may be sufficient. If not, DFT calculations can be performed, as just discussed, for additional homotops of the putative lowest-energy topologies, to more precisely approach equilibrium DFT chemical orderings.
A similar procedure could be applied to go beyond the topology approach, when a more precise energetic representation of chemical orderings is required for assessing notably higher-lying homotops. One can see from Table S1 (ESI†) a substantial energy split in one of the higher-lying topologies of 201-atomic NPs, namely by 1501 meV in Pt101Au100_594.24.0.54, but this seems to be not very often the case.
Interestingly, the energy splitting for 10 random homotops with putative lowest-energy topologies Pt70Au70_196.24.24.22 and Pt70Cu70_340.24.12.0, 1027 and 1568 meV, respectively, see structures of the homotops with minimum and maximum DFT energies of these topologies in Fig. S1 (ESI†), are notably larger than those for the 201-atom Pt–Au and Pt–Cu NPs. Again, the Pt–Cu homotops split more than the Pt–Au homotops. Our detailed analysis of this splitting reveals that not all atoms of one type (Pt–Pt, Pt–X, or X–X bonds) are equivalent, i.e. that the energy of bonds formed by Pt or X atoms to each of their first neighbours is partially dependent on the identity (and quantity) of the rest of first neighbours. This challenges the convenient assumption that all bonds between a given pair of atoms are equally strong. In addition, certain structural motifs (such as {111} facets solely composed of Au) seem to become stable for some compositions due to elusive long-range interactions neglected by the employed topological description.
Although the δ values in Table 1 may seem large, they correspond to average prediction errors 1 to 6 meV per atom. The large size of the studied particles obviously increases the errors in total energies, but changes in energy caused by permuting atoms are, on average, rather well approximated. This means that irrespective of possible inaccuracies of the present TOP approach based on the DFT structure optimization, the equilibrium chemical orderings in bimetallic nanoalloys provided by this approach approximate reasonably well those obtained with the employed DFT exchange–correlation functional. The calculated atomic-level data, which are notably more detailed than those currently accessible experimentally, are very useful for rationalizing surface reactivity of bimetallic catalysts and related applications.
Fig. 4 Equilibrium chemical orderings of ca. 4.4 nm large truncated-octahedral fcc Pt732Au731, Pt732Ag731, and Pt732Cu731 nanoparticles at 0 K with the lowest TOP energies calculated using eqn (1) for Pt70X70 particles (Pt732X731(140), left column) and Pt101X100 (particles Pt732X731(201), right column). Colors of atoms are as in Fig. 1. |
Nanoparticle | T, K | N Pt-XBOND | N XCORNER | N XEDGE | N XFACET | N XINSIDE |
---|---|---|---|---|---|---|
The total topological numbers for the Pt732X731 particles are: NBOND = 7776, NCORNER = 24, NEDGE = 108, NFACET = 440, NINSIDE = 891. | ||||||
Pt732Au731(140) | 0 | 1238 (16) | 24 (100) | 108 (100) | 440 (100) | 159 (18) |
Pt732Au731(201) | 0 | 1238 (16) | 24 (100) | 108 (100) | 440 (100) | 159 (18) |
300 | 1309 (17) | 24 (100) | 108 (100) | 439 (100) | 160 (18) | |
600 | 1502 (19) | 24 (100) | 107 (99) | 439 (100) | 161 (18) | |
1000 | 1885 (24) | 23 (96) | 108 (100) | 433 (98) | 167 (19) | |
Pt732Ag731(140) | 0 | 1238 (16) | 24 (100) | 108 (100) | 440 (100) | 159 (18) |
Pt732Ag731(201) | 0 | 1238 (16) | 24 (100) | 108 (100) | 440 (100) | 159 (18) |
300 | 1354 (17) | 24 (100) | 108 (100) | 440 (100) | 159 (18) | |
600 | 1671 (21) | 24 (100) | 107 (99) | 440 (100) | 160 (18) | |
1000 | 2129 (27) | 24 (100) | 108 (100) | 429 (98) | 170 (19) | |
Pt732Cu731(140) | 0 | 3868 (50) | 24 (100) | 48 (44) | 0 (0) | 659 (74) |
Pt732Cu731(201) | 0 | 4543 (58) | 24 (100) | 86 (80) | 57 (13) | 564 (63) |
300 | 4343 (56) | 23 (96) | 63 (58) | 58 (13) | 587 (66) | |
600 | 4244 (55) | 21 (88) | 61 (56) | 73 (17) | 576 (65) | |
1000 | 4183 (54) | 19 (79) | 55 (51) | 98 (22) | 559 (63) |
Usage of both the (140) and (201) sets of energies ε results in the same topologies of the lowest-energy orderings of Pt732Au731 and Pt732Ag731 NPs. These are core–shell orderings with Au and Ag atoms occupying all surface and a part of subsurface positions, while Pt atoms form compact cores inside the monolayer skins of coinage metals. Since Au and Ag atoms tend to surface segregate and Pt–Au and Pt–Ag bonds are destabilizing with respect to forming homometallic bonds, by 11–21 meV each, these 1238 bonds comprise only 16% of all 7776 bonds in the 1463-atom NPs, see Table 4.
The equilibrium ordering in Pt732Cu731 NP is more complex, reflecting that Pt–Cu bonds are stabilizing with respect to the corresponding homometallic bonds, and that smaller Cu atoms alloyed with Pt are less disfavored when located inside the particles compared to Au and Ag atoms. In Pt732Cu731(140) all 24 corner positions are occupied by Cu atoms, each of them stabilizing the system by −27 meV with respect to having Pt in the same positions. Occupation of almost a half of edge positions by Cu atoms alternating with Pt atoms to increase the number of favorable Pt–Cu bonds partially counterbalanced the preference, by 181 meV, of each edge Pt atom compared to having an edge Cu atom. A layered arrangement of inner Pt and Cu atoms seems to be also result from an increased number of favorable Pt–Cu bonds. No Cu atoms occupy terrace sites in line with the largest destabilization by 344 meV per terrace Cu atom with respect to having a Pt atom in those positions. Interestingly, the number of surface Cu atoms in the Pt732Cu731(201) NP, 167, is more than twice larger than in the Pt732Cu731(140) NP, 72. This finding seems counterintuitive based on the energies ε, which indicate stronger destabilization of surface Cu atoms in Pt101Cu100 NP than in Pt70Cu70 NP. It can be explained by the fact that Cu atoms in the Pt732Cu731(201) NP are scattered and mostly surrounded by Pt atoms forming more favorable Pt–Cu contacts. Indeed, there are more Pt–Cu bonds in the structure Pt732Cu731(201), 4543 (58%), than in Pt732Cu731(140), 3868 (50%), since the εPt-CuBOND values for the 201- and 140-atomic NPs are −43 and −35 meV, respectively. The energy gain due to an increased number of Pt–Cu contacts in Pt732Cu731(201) NP, −29.025 eV according to the energies ε, see Table 1, overcompensates the energy penalty caused by the presence of more surface Cu atoms, 26.429 eV. Thus, the equilibrium ordering Pt732Cu731(201) is estimated to be ca. 2.6 eV more stable than Pt732Cu731(140) one, which for such large particles translates in a difference of 1.8 meV per atom. Note that many adsorbates from the reacting media interact with Pt surface sites at least 1 eV stronger than with the corresponding Cu sites. Therefore, similar changes of the ordering patterns as those simulated for the structures Pt732Cu731(201) and Pt732Cu731(140), see Fig. 4, can be easily caused at experimental conditions in Pt–Cu NPs by adsorbing even a few reacting species from the environment.
The calculated results presented so far have corresponded to equilibrium structurers at temperature 0 K. Since the TOP method allows to approximately account for entropy effects associated with the chemical orderings (neglecting vibrational contributions), we employed a computational setup introduced elsewhere15 for estimating properties associated with Boltzmann population of different homotops. Specifically, we calculated the average topology in the Pt732X731 (X = Au, Ag, Cu) NPs in the temperature range 0–1000 K by means of the Metropolis Monte Carlo algorithm. In particular, the averaging was done over accepted configurations obtained during an MC run at a given temperature, which results in the probability of occupying of each site by either Pt or X atoms. The homotop with the most similar topology to the calculated average is then chosen as the representative model for each run. As expected, temperature increase noticeably increased the disorder in all modeled particles, see Table 4 and Fig. 5 for the results corresponding to 300, 600, and 1000 K.
Fig. 5 Equilibrium chemical orderings of Pt732Au731, Pt732Ag731, and Pt732Cu731 nanoparticles at temperatures 300 K, 600 K and 1000 K calculated using eqn (1) for the corresponding Pt101Au100, Pt101Ag100, and Pt101Cu100 particles. Colors of atoms are as in Fig. 1. |
For instance, several Pt atoms appeared at 600 and 1000 K on the surface of Pt732Au731 and Pt732Ag731 NPs that exhibited complete surface shells of Au and Ag atoms at 0 and 300 K. Concomitantly, single Au and Ag atoms migrated in the compact Pt cores. Also, the number of both Pt–Au and Pt–Ag bonds increased with the temperature increase. The fraction of Pt–Au bonds in Pt732Au731 NP increased from 16% (0 K) to 17% (300 K), 19% (600 K), and 24% (1000 K). The corresponding number of bonds in Pt732Ag731 NP increases somewhat more steep at elevated temperature, in line with less destabilizing character of Pt–Ag bonds compared with Pt–Au bonds.
In Pt732Cu731 NP the number of stabilizing Pt–Cu bonds slightly decreased at higher temperatures, from 58% at 0 K to 54% at 1000 K. Cu atoms migrate upon the temperature increase from corner and edge positions most populated at 0 K primarily to surface terrace positions.
The equilibrium chemical orderings in the Pt–Au and Pt–Ag particles with ≤201 atoms closely resemble each other: Au and Ag atoms are strongly energetically driven to be on the surface, whereas Pt atoms preferentially occupy inner positions. The immiscibility of Au and Ag atoms with Pt atoms prevents the appearance of Pt atoms on the surface of these nanoalloys as single-atom catalytic sites until the coinage metal content becomes low. Atoms in Pt–Au nanoparticles are essentially neutral, whereas in Pt–Ag nanoparticles Pt and Ag atoms are moderately charged negatively and positively, respectively. This suggests noticeably different reactivity of surface Pt sites on Pt–Ag than on single-metal Pt particles.
In Pt–Cu nanoparticles, Pt atoms are driven to the surface and most stabilized in the 9- and 8-coordinated positions of the {111} and {100} nanofacets. Pt–Cu bonds formed upon the favorable mixing of Pt and Cu atoms induce a strong Cu to Pt charge transfer. The up to half of an electron transferred to surface Pt atoms from nearby Cu atoms is expected to substantially affect the surface reactivity of both Pt and Cu sites.
Ordering trends in Pt–X nanoalloys are quite similar to the trends in previously characterized Pd–X nanoalloys, i.e. with the coinage metal atoms at the surface for X = Au or Ag, and rather well mixed orderings for X = Cu. The main difference is that heterometallic bonds slightly destabilize (with respect to homometallic bonds) Pt–Ag and Pt–Au particles, but stabilize Pd–Ag and Pd–Au ones, leading to more mixed terraces in the later two.
The most stable orderings in ca. 4.4 nm Pt732Au731 and Pt732Ag731 particles are of the Pt–core/Au(Ag)–shell type. The lowest-energy chemical ordering pattern of Pt732Cu731 particle is characterized by surface Cu atoms scattered over Pt atoms in the facets and quite randomly distributed inner atoms of both metals. Modelling the evolution of the orderings upon temperature increase revealed that heating leads to the appearance of single-atom or diatomic surface Pt sites on facets of Pt732Au731 and Pt732Ag731 particles and increases the disorder and number of surface Cu atoms on facets of Pt732Cu731 particle. New surface sites with correspondingly different chemical properties are formed upon heating.
The results presented in this work are relevant for studies in which structure and surface properties of a large number of bimetallic compositions are simulated using machine-learning and/or high-throughput approaches. Such studies often relied on models produced by directly cutting bulk structures,58,59 without considering that due to the stability of particular chemical orderings inherently different sites can be exposed on the surface.
Footnote |
† Electronic supplementary information (ESI) available: Structural and energy data calculated by DFT of the discussed in the article Pt–Au, Pt–Ag and Pt–Cu nanoparticles (PDF). See DOI: 10.1039/d1ma00529d |
This journal is © The Royal Society of Chemistry 2021 |