ACS Publications. Most Trusted. Most Cited. Most Read
My Activity
CONTENT TYPES

Figure 1Loading Img

(Meta-)stability and Core–Shell Dynamics of Gold Nanoclusters at Finite Temperature

Cite this: J. Phys. Chem. Lett. 2019, 10, 3, 685–692
Publication Date (Web):January 25, 2019
https://doi.org/10.1021/acs.jpclett.8b03397

Copyright © 2019 American Chemical Society. This publication is licensed under CC-BY.

  • Open Access

Article Views

1452

Altmetric

-

Citations

LEARN ABOUT THESE METRICS
PDF (4 MB)
Supporting Info (1)»

Abstract

Gold nanoclusters have been the focus of numerous computational studies, but an atomistic understanding of their structural and dynamical properties at finite temperature is far from satisfactory. To address this deficiency, we investigate gold nanoclusters via ab initio molecular dynamics, in a range of sizes where a core–shell morphology is observed. We analyze their structure and dynamics using state-of-the-art techniques, including unsupervised machine-learning nonlinear dimensionality reduction (sketch-map) for describing the similarities and differences among the range of sampled configurations. In the examined temperature range between 300 and 600 K, we find that whereas the gold nanoclusters exhibit continuous structural rearrangement, they are not amorphous. Instead, they clearly show persistent motifs: a cationic core of 1–5 atoms is loosely bound to a shell which typically displays a substructure resulting from the competition between locally spherical versus planar fragments. Besides illuminating the properties of core–shell gold nanoclusters, the present study proposes a set of useful tools for understanding their nature in operando.

Transition-metal nanoclusters have been the focus of a large number of experimental and computational studies in physics and chemistry because of their expected technological applications in diverse areas, including catalysis, (1) optics, (2) and biomedicine. (3) Their physical and chemical properties are dependent on many characteristics such as size, (4) shape, (5) charge state, (6) ligand effects, (7) and temperature. (8) Transition-metal nanoclusters typically exhibit catalytic activity at high temperature, (9) but gold nanoclusters have recently received special attention because of the discovery of their catalytic activity at low temperature. (10)

Experimental and computational studies of both colloidal (11−15) and bare (16−23) gold nanoclusters have achieved important advances in the last decades. Most of the reported ab initio studies have been dedicated to searching for their putative global (energy) minimum configurations (pGMC) in the gas phase at zero temperature. (14) An empirical-potential molecular dynamics (MD) study of Au75, Au146, and Au457 showed that their melting temperatures were well below the melting temperature of bulk gold (1090 K, at standard pressure) but also that structural solid-to-solid transitions were present well below their melting temperatures. (24) For instance, for Au75, the melting temperature is around 550 K, but around 350 K there is a structural transition from truncated-decahedral to icosahedral.

The present study focuses on the gas-phase gold nanoclusters Au25, Au38, and Au40. Au25 was chosen because it is expected to be the smallest cluster that has a core, in this case a single Au atom, in the low-lying isomer population. (18) Au38 and Au40 were chosen because they are both expected to have tetrahedral Au4 cores and low-symmetry structures have been proposed as their pGMCs. (22,23) We use ab initio Born–Oppenheimer MD, an important computational technique that allows for the sampling of the configurational space of systems with many atoms, together with yielding reliable information on the dynamics of the system, for example, in terms of time-correlation functions. When the Born–Oppenheimer MD equations of motion are extended by adding a thermostat, the canonical ensemble (at temperature T > 0 K) is sampled. In the specific case of density-functional theory (DFT)-based MD, the energy of the system and forces between atoms are determined by solving the Kohn–Sham equations. Hence, the ground-state electronic structure is accessible along the simulated trajectories.

DFT-based MD studies on small Au clusters have been performed in the past, aimed at describing the thermal evolution of the ground-state structures and the competition between structural isomers at a given temperature, (25−27) as well as at accessing the infrared (vibrational) spectra for comparison with experiments. (28) We analyze DFT-based MD trajectories of Au25, Au38, and Au40 at room and higher temperatures in order to address important questions regarding the structural and dynamical aspects specific to this range of cluster sizes: (i) Is the core–shell morphology persistent at finite temperature? (ii) Do the atoms constituting the core exchange their positions with those constituting the shell? (iii) Does the shell of different isomers have recurrent structural motifs?

We ran DFT-based MD within the Perdew–Burke–Ernzerhof (29) (PBE) approximation for the electronic exchange–correlation functional, corrected for the many-body dispersion (MBD) interactions, (30) in the all-electron full-potential implementation of the FHI-aims package. (31) The time step for the numerical integration of the MD equations of motion was set to 10 fs, and we applied a Gaussian broadening of 1 meV for the occupation at the Fermi level. Canonical sampling was imposed via the stochastic velocity-rescaling thermostat. (32) Further MD details are described in the Supporting Information. These settings have been already adopted in refs (27,28,and33) and were shown to yield observable quantities in good agreement with experiments in the size range between 3 and 20 gold atoms. Gold nanoclusters are strongly influenced by relativistic effects, (34,35) as the contraction of the 6s orbital and expansion of the 5d results in a smaller s–d gap and an increased directionality in the Au–Au bonding. In the present study, relativistic effects are included via the “atomic” zeroth-order regular approximation to the Dirac equation. (36)

The initial configurations for the MD trajectory were (Figure 1) the pGMC structures described in ref (19) for Au38 and Au40, both containing a tetrahedral Au4 core, and a higher-energy isomer for Au40 (also from ref (19)), containing a Au3 core. Furthermore, to increase diversity in the starting point of the MD trajectories of Au38 and Au40, we also selected the respective pGMC of a potential-energy surface described through the embedded-atom method (EAM), globally optimized via the revised basin-hopping Monte Carlo (RBHMC) algorithm, (37) both containing a Au5 core. For Au25, we adopted the embedded-atom-RBHMC pGMC, which has a one-atom core. More details on the RBHMC search are given in Figure S1 of the Supporting Information. We note that close-packed structures, such as icosahedra and decahedra, which can be important for larger gold nanoclusters, are not relevant (i.e., energetically very unfavored) for the smaller sizes considered in the present Letter. (38) For each starting point, one or more ab initio MD trajectories of 100 ps is run, where the first interval of 50 ps is thermostated at a temperature between 400 and 600 K (see Table S1 for details on the single trajectories) and the last interval of 50 ps is at “room temperature”, i.e., 300 K.

Figure 1

Figure 1. Relative total energies and configurations used as starting points of the MD trajectories. Structures labeled as Au38J, Au40J1, and Au40J2 are from ref (19), while structures labeled as Au25, Au38, and Au40 are found via EAM-RBHMC and further optimized with PBE+MBD (see text for details). The core (shell) atoms are indicated by blue (yellow). The labeling Auc@Aus marks the number of atoms in the core (c) and shell (s).

As detailed below, the statistical analysis of the structure and dynamics of the studied clusters is performed by means of radial-distribution functions, bond-lifetime analysis, nonlinear dimensionality reduction (in order to represent, by exploiting “structure similarities”, the high-dimensional configurational space in two dimensions), and state-of-the-art Boltzmann reweighting (in order to use the MD sampling at all temperatures in the most efficient way).

The (center-of-mass centered) radial-distribution function g(r) is the histogram in which bin i counts how many atoms are found between distance ri–1 and ri from the center of mass of the cluster, averaged over all configurations along a trajectory. Here, r0 = 0 and the radial increment ri+1ri = δr is constant and set to 0.1 Å. For extended systems, the histogram is typically normalized by the volume of the spherical shell between ri+1 and ri, but for finite systems, this normalization is not relevant and omitted here. This descriptor, invariant with respect to rigid rotations and translations of the whole cluster and to permutations of atom identities, gives direct and statistical access to the radially layered structure of the clusters and allows for a parameter-free definition of the core and shell regions. Figure 2a shows representative g(r) plots for one trajectory of Au38 and one trajectory of Au40. At both cluster sizes, three clear regions can be distinguished. Some gold atoms are found in the range 1.0–2.5 Å. These are the atoms constituting the core. Another group of atoms is found in the range 2.5–5.5 Å. These are the atoms constituting the shell. A third group of atoms appears at larger distances. A minimum in g(r) at r ≈ 5.1 Å (Au38) and r ≈ 5.5 Å (Au40) suggests that atoms contributing to the outer part of the distribution form a well-separated group from the shell atoms. Visual inspection of the cluster structures (e.g., green-colored spheres in Figure 4) reveals that these are lower-coordinated atoms “protruding” from an otherwise triangular-lattice shell, the latter typically composed of 6–7 fold coordinated atoms. We therefore name this outer region protrusions. The g(r) plots for all trajectories, classified also by temperature, are shown in Figures S2–S4. For Au38 and Au40, the three-region feature is observed for all trajectories and temperatures, with cutoff radii at 2.5 Å for core–shell and in the range of 5.0–5.5 Å for shell-protrusions. The average numbers of gold atoms in the core/shell/protrusion regions (obtained by integrating the g(r) in panels a and d, which is the same as taking the time average of the trajectories in panels c and f) are, for Au38 at 300 K (600), 4.0/31.0/3.0 (4.1/30.2/3.7); for Au40 at 300 K (600), 4.0/31.8/4.2 (4.1/32.2/3.6). We notice that for Au38, atoms move on average from protrusion to shell region when going from 600 to 300 K, while the opposite trend is observed for Au40. The numbers of atoms in the core tends to remain constant. These temperature effects are accompanied by a less clear boundary between shell and protrusion regions for both cluster sizes at 600 K, compared to 300 K, signaled by a less marked minimum of g(r) at r ≈ 5.5 Å. Another observation, from the trajectories in panels c and f, is that both shell and protrusion regions for Au38 are rather stiff at 300 K while more dynamical at 600 K and for Au40 at both temperatures. For Au25, only a (one-atom) core region and a shell region (the latter in the range of 2.5–5.5 Å, as for the larger clusters) is observed.

Figure 2

Figure 2. (a) Radial distribution function, g(r), and cutoff distances (solid vertical lines) with respect to the cluster’s center of mass defining the core, shell, and protrusion regions for the Au38 nanocluster at 600 and 300 K. (b) Evolution of the Hirshfeld charges for core and shell atoms and (c) evolution of number of atoms belonging to protrusions (green line), shell (orange line), and core (blue line) regions at 600 and 300 K for Au38. Panels d–f show the same quantities as in panels a–c, respectively, but for Au40.

We note here that atoms are observed to change region along the MD trajectories, so their assignment to a region is dynamical and based on solely the distance from the center of mass. This flexible definition is crucial for our subsequent analysis. In Figure 2, we also show the time evolution of the Hirshfeld charges (39) of the core and shell+protrusion regions (panels b and e), and of the number of atoms belonging to these regions (panels c and f), for Au38 and Au40. These observables are reported for all trajectories in Figures S2–S5. The qualitative behavior is remarkably consistent at all cluster sizes and temperatures: the core is always fractionally cationic, while the shell+protrusions is always fractionally anionic.

The number of atoms in the core is very stable. For Au25, it is always one, and for Au38, it is almost always four, with infrequent fluctuation to five (in the trajectories that started from the structure with a five-atom core, one atom quickly, in a few picoseconds, moves to the shell region). For Au40, the behavior is more complex. Trajectories starting with four atoms in the core maintain a four-atom core (with sporadic fluctuations to five) throughout the time evolution, even at the higher temperature (600 K). However, the trajectory starting from a three-atom core (structure “J2” in Figure 1) fluctuates between three and four atoms in the core for the 50 ps at 600 K and remains with a three-atom core during the subsequent 50 ps time interval at 300 K, indicating a kinetic trapping in a metastable state. When starting with a five-atom core (EAM-RBHMC pGMC structure), it takes 500 K and more than 10 ps to see the trigonal-bipyramidal five-atom core transform into a tetrahedral four-atom core. At 600 K, the transformation is faster (fraction of a picosecond, but followed by frequent fluctuation back to a five-atom core), while at 400 K, the five-atom core remains stable. The subsequent parts of the trajectories at 300 K maintain the number of atoms in the core, as at the higher temperature. In this case, we have also the signature of a kinetic trapping that is easily overcome at 600 K.

The number of atoms belonging to either shell or protrusions regions (for Au38 and Au40) is quite stable, but with larger fluctuations (±2–3 atoms around the average value along the trajectories). Average sizes of the protrusions regions range from 2 to 7 atoms, depending on the cluster size and the temperature.

To further analyze the nature of the different regions in the clusters, we measured the average bond lifetimes, according to whether the bonded pair belongs to one region or is across regions. To this purpose, we first define a bonded pair (at a given time step) as any pair of atoms at a distance smaller than 3.6 Å. This cutoff value is assigned as the first minimum of the pair-distribution function. It is a concept similar to the g(r) defined above, but here the distance between any pair of atoms is binned. The position of the first minimum (see Figure S6) is relatively independent of cluster size, temperature, and region to which the atoms belong. Next, if a bond is found at any time step, it is monitored until it is first broken (i.e., until the distance between the two atoms exceeds the threshold): The elapsed time is the bond lifetime, and it is binned considering whether both atoms are in the core, or one in the core and one in the shell, etc. In Figure 3, we show τ, the region-dependent average value of the bond lifetime, at each cluster size and temperature (if more trajectories sampled the same temperature, we averaged over them). We also show the sample standard deviation of each bond-lifetime histogram as an error bar in order to quantify the spread of this quantity under the different conditions. More details on the bond lifetimes, broken down to single trajectories, are given in Table S2. We observe a striking difference in behavior between bonds shared by atoms both belonging to the core, or both belonging to the shell, and bonds across core and shell. The latter are systematically at least 3–4 times shorter lived than the former. The shell–protrusion bonds behave similarly to shell–shell bonds, suggesting that the two regions are structurally separated but dynamically very similar. Protrusion–protrusion bonds were too few to yield robust statistics at all sizes and temperatures. The dynamics of the core–shell bonds suggests a picture of a shell that is loosely bound to the core. Visual inspection of the trajectories of Au40 and Au38 confirms that the core atoms behave like a inner cluster “confined”, and almost freely rotating, inside the shell. In the case of Au25, the single-atom core “rattles” around the center of mass. Consistently, g(r) for Au25 (Figure S3) shows that the one-atom core is positioned on average at about 0.25 Å from the center of mass.

Figure 3

Figure 3. (a) Average bond lifetime, τ, and sample standard deviation (bars) among the core–core atoms (τcc), core–shell atoms (τcs), shell–shell atoms (τss), and shell-protrusions (τps) atoms, for the Au25, Au38, and Au40 nanoclusters as a function of the sampled temperatures. The lines are a guide to the eye.

In order to describe the statistical distribution of structures at each cluster size, as sampled along the MD trajectories, we adopted an unsupervised machine-learning approach, namely, the nonlinear dimensionality-reduction algorithm termed sketch-map, as introduced in ref (40). First, similarly to ref (41), we describe each configuration along a MD trajectory with a coordination histogram, where the jth bin is the number of atoms that have coordination j, i.e., that have j atoms at a distance within the same threshold defined above for the bond breaking. A graphical example of how a cluster structure is mapped into a coordination histogram is given in Figure S7. The histogram is normalized; that is, each bin is divided by the number of atoms in the cluster. For the clusters considered, we found that the maximum coordination is 12, and each histogram was therefore set to have dimension 13, starting from 0-fold coordinated atoms (i.e., atoms evaporated from the clusters) which were, however, not observed in the present study. This 13-dimensional representation is already a dimensionality reduction from the 3N coordinates for an N-atom cluster. Such representation is roto-translational and permutation invariant, similar to g(r). In this preliminary dimensionality reduction, some information is necessarily lost, as two different clusters may accidentally have the same coordination histogram. Nonetheless, the coordination histogram captures essential information regarding the bonding network for the clusters. Next, we define a distance in the coordination-histogram representation, which is simply the Euclidean distance between histograms, considered as vectors. The sketch-map algorithm (40) then embeds, in two dimensions, the high-dimensional points (here, in the coordination-histogram representation), by representing points that are close (distant) in high dimension as close (distant) in the two-dimensional (2-D) map. The axes in the plots do not have any physical meaning and cannot even be labeled explicitly; what matters are the distances. Still, each high-dimensional point is mapped into a specific point on the 2-D map. Therefore, by discretizing the 2-D coordinates, one can count how many points land in each discrete (2-D) bin and color-code the map according to the number of points in each bin. For better statistics, we adopted a Boltzmann reweighting approach, the multistate Bennet acceptance-ratio (MBAR) algorithm, (42) to estimate the probability that a structure falls in a given bin, by using the trajectories at all temperatures, while the shown plot refers to 300 K only. In this way, the above-described kinetic-trapping issues at 300 K are mitigated by simultaneous use of trajectories at higher temperatures, where no kinetic trapping is observed. For a converged sampling, the logarithm of these probabilities is proportional to the negative of the free-energy of the given AuN cluster. In our case, however, the samplings are most likely not converged, and the probabilities are therefore only qualitative. The results for Au38 and Au40 are shown in Figure 4 (results for Au25 are in Figure S8). Absolute errors for the same sketch-maps, also evaluated via MBAR, are in Figures S9–S11.

Figure 4

Figure 4. Probability of cluster configurations at 300 K, calculated via MBAR as a function of the sketch-map coordinates for the population sampled for Au38 (left side, based on a total of 400 ps of MD sampling) and Au40 (right side, based on 500 ps MD). The selected structures distinguish core (blue), shell (yellow), and protrusion (green) regions. For each selected structure, also the corresponding 13-dimensional coordination-histogram, the input for the sketch-map dimensionality reduction, is shown. The dots mark structures selected along the MD runs, positioned in the areas of highest probability, and the stars locate the corresponding geometrically relaxed structures (dot/star of the same color indicate initial/final structures for a geometry relaxation). For Au38, the pink star is the pGMC found by us (see its structure in Figure S15). The pGMC identified in ref (19) is marked by a diamond. For Au40, the red star is the pGMC found by us, which is also the pGMC identified in ref (19).

Because the positioning of the data points on the sketch-maps is determined only by their 13-dimensional coordination-histogram representation, the criterion of the color coding can be varied at fixed sketch-map, to highlight different aspects of the phase-space sampling. In Figures S12–S14, we assign one color in the sketch-maps for data points belonging to the same MD trajectory, to show that the trajectories all diffuse over broad regions of the sketch-map and therefore of the configurational space.

For Au38 (Figure 4, left), we observe three rather distinct configurational basins. The structures belonging to the left and center basins have a majority of 7-fold coordinated atoms, while for the right basin, structures have a similarly high population of 6- and 7-fold coordinated atoms. The difference between the left and center basins is that structures in the center basin are more spherically shaped, while structures in the left basin are more distorted, with locally planar portions of the shell. Incidentally, both the RBHMC and “J” initial structures belong to the center basin. Furthermore, structures similar to the J structure are visited during the MD trajectories started from the RBHMC structure, suggesting a rather broad sampling of the configurational space along the MD trajectories. The connections between the three basins are narrow funnels, which means that, during the MD trajectories, while structures diffuse freely inside the basins, they need specific, coordinated, and infrequent structural rearrangements in order to pass from one basin to another. The power of the sketch-map is to reveal this latent order, invisible to human eye in the high-dimensional representation, even though there might not be a simple description in terms of geometrical features of the structures.

For Au40 (Figure 4, right), we also observe three basins, albeit less well-defined than for Au38. The center-bottom basin contains structures with a trigonal-bipyramidal five-atom core and a preference for 6- and 7-fold coordinated atoms. The left basin has structures with a tetrahedral four-atom core, while in the right basins there is a combination of structures with three- and four-atom cores, where the common structural motif (of the whole cluster) is the twisted trigonal pyramid. Structures in both the left- and right basins have a majority of 7-fold coordinated atoms.

For Au25 (see Figure S6), we found tubular-like structures (always with one atom in the core) with one side distorted, an aspect reminiscent of the known Au24 pGMC. (20) In a different basin, Au25 shows a further distortion of the tubular structure, yielding locally planar motifs.

The sketch-maps in Figure 4 (and Figure S6) report also the position of the pGMC and other local minima. We note that these structures are not located in the main basins, which means that these structures are not in the regions of the configurational space mostly visited by the MD sampling. Keeping in mind the caveats expressed above, we refer to these “darker”, most visited regions as low-free-energy regions in the following. To rationalize the finding that low-energy, relaxed structures deviate significantly from the low-free-energy configurations, we selected from the MD trajectories several structures in the lowest-free-energy areas of the sketch-maps and geometrically relaxed them to find their local minimum energy structures. This is shown in Figure 4 (and Figure S6) with differently colored dots (selected structures from the MD sampling) and stars (relaxed structures), where dot–star pairs with the same color depict initial and final structures for one relaxation. In all cases, the relaxed structures are in higher-free-energy regions compared to the selected initial structures. Interestingly, one of the selected structures from the high-probability regions, labeled as 40d, relaxes into the pGMC, further supporting the observation that high-probability structures at finite temperature are significantly different, based on their coordination histograms, from the pGMC as well as from the other lower-energy local minima. We inspected the change upon relaxation of (i) gyration radius, (ii) number of core–shell bonds, and (iii) the average effective coordination number (43,44) (ECN). The ECN is an indicator of the average coordination of atoms in a cluster. It adopts a self-consistent, parameter-free, smoothly decaying definition of bond and therefore coordination number for each atom. These values are reported in Table S3. As a general trend, the gyration radius decreases and the number of core–shell bonds increases upon relaxation. At the same time, the ECN always increases. This indicates that there is a contraction of the clusters upon relaxation and, in particular, a tightening of the core–shell binding. Together with the bond lifetime analysis (Figure 3), this observation further strengthens the picture of typical gold cluster structures at 300 K and higher temperatures with weak coupling between core and shell and therefore larger configurational freedom (hence lower free energy) than relaxed structures.

An analysis of the dynamics of the single trajectories projected onto the 2-D maps reveals that each trajectory of Au38 and Au40 visits along its 50 ps all the main basins in the map (i.e., those with an exemplary structure shown in the plots), swapping between neighboring basins between 10 and 100 times. Much slower is the diffusion for Au25, where at 300 K the trajectories never leave the basin they start in, but in all three cases, during the initial 50 ps at higher temperature, the trajectory swaps few times between the two main basins. We can therefore conclude that, although full ergodicity of our trajectories cannot be ensured and indeed we notice signatures of some kinetic trapping, single trajectories do not merely (anharmonically) vibrate in one basin but their diffusion across the 2-D maps is significant.

We note, incidentally, that the relaxation of the low-free-energy structures let us identify pGMC for Au25 and Au38, which are previously unreported in the literature (pink stars in Figures 4 (left) and S6). The corresponding structures are shown in Figure S15. For Au40, the pGMC we found (red star in Figure 4 (right)) corresponds with the pGMC in ref (19).

Finally, we investigated the reason for the stability of three-dimensional four- and five-atom cores, while isolated neutral Au4 and Au5 are known to be planar. (28) We measured the amount of positive charge needed to stabilize the 3-D structure versus the planar one for both Au4 and Au5. In Figure 5, we show the energy of the 3-D structure, relative to the planar one for both Au4 and Au5, where both structures are geometrically unrelaxed for each charge. Some structures like the neutral tetrahedron or trigonal-bipyramidal are mechanically unstable, i.e., have imaginary vibrational frequencies (see Figures S16 and S17). In Figure S16, we show the relaxation of the structures by fixing their symmetry and in Figure S17 without fixing symmetry. We find that tetrahedral Au4 becomes more stable than planar Au4 with a charge of 1.5e, while trigonal-bipyramidal Au5 needs 2.0e to become more stable than planar Au5. These charge values, even without symmetry constraints, are much larger than the small positive charges observed for the core atoms. Therefore, charge alone is not sufficient to stabilize the 3-D structures, leaving confinement and bonding interactions with the shell as likely reasons for the stabilization.

Figure 5

Figure 5. Left (right): Relative stability of the tetrahedral (trigonal-bipyramidal) structure with respect to the planar Au4 (Au5) structure, as a function of the positive total charge of the cluster.

In summary, we have presented a theoretical study of the stability and dynamics of Au25, Au38, and Au40 nanoclusters whose configurational space is sampled via Born–Oppenheimer molecular dynamics at various temperatures. We showed that in the considered range of temperatures (between 300 and 600 K) the gold clusters are not amorphous. Instead, they exhibit a dynamical core–shell structure, with a cationic core (single Au for Au25 and mainly tetrahedral Au4 for Au38 and Au40) loosely bound to an outer anionic flexible shell. We also identify, as a persistent feature, a substructure in the cluster shells- one triangular-lattice region with 6- and 7-fold coordinated atoms and an outer region with fewer lower-coordinated atoms that we term protrusions. By using sketch-maps—two-dimensional embeddings of the coordination-histogram representation of the clusters—and Boltzmann reweighting, we built approximate free-energy landscapes, at room temperature, of the studied nanoclusters. These show that the global and local minimum energy structures differ significantly from the structures in low free-energy areas, i.e., the most probable at finite temperature. Obtained by using a suite of statistical tools that can be applied readily to similar systems, our results are able to highlight some typical characteristics of the finite-temperature populations of Au25, Au38, and Au40, providing insights into the stability and dynamics of their core–shell regions. Our results (especially those for Au38 and Au40) suggest a possible mechanism of (equilibrium) cluster growth. If gold atoms are added to a nanocluster in the size range that we have considered, they will tend to remain in the shell and protrusion regions until a critical value is reached when one of the gold atoms stably transfers to the core.

Supporting Information

ARTICLE SECTIONS
Jump To

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.8b03397.

  • Detailed computational details for the EAM-RBHMC, DFT-MD, and sketch-map-training calculations; the g(r) and Hirshfeld analyses for all MD trajectories; the resulting sketch-map for Au25; MBAR-estimated error bars on the probabilities reported in all the sketch-maps; detailed bond-lifetime analysis for each trajectory; detailed information on all found local minima; and further information on stability analysis of charged Au4 and Au5 (PDF)

Terms & Conditions

Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

Author Information

ARTICLE SECTIONS
Jump To

Acknowledgments

ARTICLE SECTIONS
Jump To

We thank Matthias Scheffler for his support and Bryan Goldsmith for a critical reading of the manuscript and insightful suggestions. D.G.-S. and J.L.F.D.S. acknowledge support from FAPESP (São Paulo Research Foundation, Grant Numbers 2017/11631-2 and 2014/22044-2), Shell, and the strategic importance of the support given by ANP (Brazils National Oil, Natural Gas and Biofuels Agency) through the R&D levy regulation. D.G.-S. and J.L.F.D.S. also thank the National Council for Scientific and Technological Development (CNPq) and the Coordination for Improvement of Higher Level Education (CAPES) for financial support. I.H. thanks the National Sciences and Engineering Research Council (NSERC) for financial support. L.M.G. acknowledges for financial support the European Union Horizon 2020 research and innovation program (#676580: The NOMAD Laboratory—an European Center of Excellence and #740233: TEC 1p) and the Berlin Big-Data Center (BBDC, #01IS14013E). Computer time was provided by the Max Planck Computing and Data Facility, the Laboratory of Advanced Scientific Computing (University of São Paulo), the Department of Information Technology—Campus São Carlos, and Compute Canada.

References

ARTICLE SECTIONS
Jump To

This article references 44 other publications.

  1. 1
    III Aiken, J. D.; Finke, R. G. A. A Review of Modern Transition-Metal Nanoclusters: Their Synthesis, Characterization, and Applications in Catalysis. J. Mol. Catal. A: Chem. 1999, 145, 144,  DOI: 10.1016/S1381-1169(99)00098-9
  2. 2
    Jin, R. Atomically Precise Metal Nanoclusters: Stable Sizes and Optical Properties. Nanoscale 2015, 7, 15491565,  DOI: 10.1039/C4NR05794E
  3. 3
    Marjomäki, V.; Lahtinen, T.; Martikainen, M.; Koivisto, J.; Malola, S.; Salorinne, K.; Pettersson, M.; Häkkinen, H. Site-Specific Targeting of Enterovirus Capsid by Functionalized Monodisperse Gold Nanoclusters. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 12771281,  DOI: 10.1073/pnas.1310973111
  4. 4
    Chaves, A. S.; Piotrowski, M. J.; Da Silva, J. L. F. Structural, Energetics, and Electronic Properties Evolution of the 3d, 4d, and 5d Transition-Metal Clusters (30 TMn Systems for n= 2 – 15): A Density Functional Theory Investigation. Phys. Chem. Chem. Phys. 2017, 19, 1548415502,  DOI: 10.1039/C7CP02240A
  5. 5
    Mostafa, S.; Behafarid, F.; Croy, J. R.; Ono, L. K.; Li, L.; Yang, J. C.; Frenkel, A. I.; Cuenya, B. R. Shape-Dependent Catalytic Properties of Pt Nanoparticles. J. Am. Chem. Soc. 2010, 132, 1571415719,  DOI: 10.1021/ja106679z
  6. 6
    Chaves, A. S.; Rondina, G. G.; Piotrowski, M. J.; Tereshchuk, P.; Da Silva, J. L. F. The Role of Charge States in the Atomic Structure of Cun and Ptn (n = 2 – 14 Atoms) Clusters: A DFT Investigation. J. Phys. Chem. A 2014, 118, 1081310821,  DOI: 10.1021/jp508220h
  7. 7
    Jadzinsky, P. D.; Calero, G.; Ackerson, C. J.; Bushnell, D. A.; Kornberg, R. D. Structure of a Thiol Monolayer-Protected Gold Nanoparticle at 1.1 Å Resolution. Science 2007, 318, 430433,  DOI: 10.1126/science.1148624
  8. 8
    Yamauchi, H.; Ito, S.; Yoshida, K.-i.; Itoh, T.; Tsuboi, Y.; Kitamura, N.; Miyasaka, H. Temperature Near Gold Nanoparticles under Photoexcitation: Evaluation Using a Fluorescence Correlation Technique. J. Phys. Chem. C 2013, 117, 83888396,  DOI: 10.1021/jp311173j
  9. 9
    Sugimura, Y.; Suzuki, Y. A. A High-Temperature Catalytic Oxidation Method for the Determination of Non-Volatile Dissolved Organic Carbon in Seawater by Direct Injection of a Liquid Sample. Mar. Chem. 1988, 24, 105131,  DOI: 10.1016/0304-4203(88)90043-6
  10. 10
    Haruta, M. When Gold Is Not Noble: Catalysis by Nanoparticles. Chem. Rec. 2003, 3, 7587,  DOI: 10.1002/tcr.10053
  11. 11
    Häkkinen, H.; Barnett, R. N.; Landman, U. Electronic Structure of Passivated Au38 (SCH3)24 Nanocrystal. Phys. Rev. Lett. 1999, 82, 32643267,  DOI: 10.1103/PhysRevLett.82.3264
  12. 12
    Heaven, M. W.; Dass, A.; White, P. S.; Holt, K. M.; Murray, R. W. Crystal Structure of the Gold Nanoparticle [N(C8 H17)4][Au25 (SCH2 CH2 Ph)18]. J. Am. Chem. Soc. 2008, 130, 37543755,  DOI: 10.1021/ja800561b
  13. 13
    Farrag, M.; Tschurl, M.; Dass, A.; Heiz, U. Infra-Red Spectroscopy of Size Selected Au25, Au38 and Au144 Ligand Protected Gold Clusters. Phys. Chem. Chem. Phys. 2013, 15, 12539,  DOI: 10.1039/c3cp51406d
  14. 14
    Azubel, M.; Koivisto, J.; Malola, S.; Bushnell, D.; Hura, G. L.; Koh, A. L.; Tsunoyama, H.; Tsukuda, T.; Pettersson, M.; Häkkinen, H.; Kornberg, R. D. Electron Microscopy of Gold Nanoparticles at Atomic Resolution. Science 2014, 345, 909912,  DOI: 10.1126/science.1251959
  15. 15
    Jin, R.; Zeng, C.; Zhou, M.; Chen, Y. Atomically Precise Colloidal Metal Nanoclusters and Nanoparticles: Fundamentals and Opportunities. Chem. Rev. 2016, 116, 1034610413,  DOI: 10.1021/acs.chemrev.5b00703
  16. 16
    Garzón, I. L.; Beltrán, M. R.; González, G.; Gutierrez-González, I.; Michaelian, K.; Reyes-Nava, J. A.; Rodriguez-Hernández, J. I. Chirality, defects, and disorder in gold clusters. Eur. Phys. J. D 2003, 24, 105109,  DOI: 10.1140/epjd/e2003-00187-4
  17. 17
    Xing, X.; Yoon, B.; Landman, U.; Parks, J. H. Structural Evolution of Au Nanoclusters: From Planar to Cage to Tubular Motifs. Phys. Rev. B: Condens. Matter Mater. Phys. 2006, 74, 165423,  DOI: 10.1103/PhysRevB.74.165423
  18. 18
    Bulusu, S.; Li, X.; Wang, L.-S.; Zeng, X. C. Structural Transitions from Pyramidal to Fused Planar to Tubular to Core/Shell Compact in Gold Clusters: Aun (n= 21–25). J. Phys. Chem. C 2007, 111, 41904198,  DOI: 10.1021/jp068897v
  19. 19
    Jiang, D.; Walter, M. Au40: A Large Tetrahedral Magic Cluster. Phys. Rev. B: Condens. Matter Mater. Phys. 2011, 84, 193402,  DOI: 10.1103/PhysRevB.84.193402
  20. 20
    Wang, L.; Wang, L. Probing the Electronic Properties and Structural Evolution of Anionic Gold Clusters in the Gas Phase. Nanoscale 2012, 4, 4038,  DOI: 10.1039/c2nr30186e
  21. 21
    Zeng, C.; Chen, Y.; Das, A.; Jin, R. Transformation Chemistry of Gold Nanoclusters: From One Stable Size to Another. J. Phys. Chem. Lett. 2015, 6, 29762986,  DOI: 10.1021/acs.jpclett.5b01150
  22. 22
    Shao, N.; Huang, W.; Mei, W.-N.; Wang, L. S.; Wu, Q.; Zeng, X. C. Structural Evolution of Medium-Sized Gold Clusters Aun–(n = 36, 37, 38): Appearance of Bulk-Like Face Centered Cubic Fragment. J. Phys. Chem. C 2014, 118, 68876892,  DOI: 10.1021/jp500582t
  23. 23
    Pande, S.; Huang, W.; Shao, N.; Wang, L.; Khetrapal, N.; Mei, W.; Jian, T.; Wang, L.; Zeng, X. C. Structural Evolution of Core–Shell Gold Nanoclusters: Aun (n = 42 – 50). ACS Nano 2016, 10, 1001310022,  DOI: 10.1021/acsnano.6b04330
  24. 24
    Cleveland, C. L.; Luedtke, W. D.; Landman, U. Melting of Gold Clusters. Phys. Rev. B: Condens. Matter Mater. Phys. 1999, 60, 50655077,  DOI: 10.1103/PhysRevB.60.5065
  25. 25
    de Bas, B. S.; Ford, M. J.; Cortie, M. B. Melting in Small Gold Clusters: a Density Functional Molecular Dynamics Study. J. Phys.: Condens. Matter 2006, 18, 5574,  DOI: 10.1088/0953-8984/18/1/004
  26. 26
    Kang, J.; Wei, S.; Kim, Y. Origin of the Diverse Melting Behaviors of Intermediate-Size Nanoclusters: Theoretical Study of AlN (N = 51–58, 64). J. Am. Chem. Soc. 2010, 132, 1828718291,  DOI: 10.1021/ja107683m
  27. 27
    Beret, E. C.; Ghiringhelli, L. M.; Scheffler, M. Free Gold Clusters: Beyond the Static, Monostructure Description. Faraday Discuss. 2011, 152, 153,  DOI: 10.1039/c1fd00027f
  28. 28
    Ghiringhelli, L. M.; Gruene, P.; Lyon, J. T.; Rayner, D. M.; Meijer, G.; Fielicke, A.; Scheffler, M. Not so Loosely Bound Rare Gas Atoms: Finite-Temperature Vibrational Fingerprints of Neutral Gold-Cluster Complexes. New J. Phys. 2013, 15, 083003  DOI: 10.1088/1367-2630/15/8/083003
  29. 29
    Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 38653868,  DOI: 10.1103/PhysRevLett.77.3865
  30. 30
    Tkatchenko, A.; Ambrosetti, A.; DiStasio, R. A. Interatomic Methods for the Dispersion Energy Derived from the Adiabatic Connection Fluctuation-Dissipation Theorem. J. Chem. Phys. 2013, 138, 074106  DOI: 10.1063/1.4789814
  31. 31
    Blum, V.; Gehrke, R.; Hanke, F.; Havu, P.; Havu, V.; Ren, X.; Reuter, K.; Scheffler, M. Ab Initio Molecular Simulations with Numeric Atom-Centered Orbitals. Comput. Phys. Commun. 2009, 180, 21752196,  DOI: 10.1016/j.cpc.2009.06.022
  32. 32
    Bussi, G.; Donadio, D.; Parrinello, M. Canonical Sampling Through Velocity Rescaling. J. Chem. Phys. 2007, 126, 014101  DOI: 10.1063/1.2408420
  33. 33
    Yu, C.; Harbich, W.; Sementa, L.; Ghiringhelli, L. M.; Aprá, E.; Stener, M.; Fortunelli, A.; Brune, H. Intense Fluorescence of Au20. J. Chem. Phys. 2017, 147, 074301  DOI: 10.1063/1.4996687
  34. 34
    Häkkinen, H.; Moseler, M.; Kostko, O.; Morgner, N.; Hoffmann, M. A.; v. Issendorff, B. Symmetry and Electronic Structure of Noble-Metal Nanoparticles and the Role of Relativity. Phys. Rev. Lett. 2004, 93, 093401,  DOI: 10.1103/PhysRevLett.93.093401
  35. 35
    Huang, W.; Ji, M.; Dong, C.-D.; Gu, X.; Wang, L.-M.; Gong, X. G.; Wang, L. Relativistic Effects and the Unique Low-Symmetry Structures of Gold Nanoclusters. ACS Nano 2008, 2, 897904,  DOI: 10.1021/nn800074b
  36. 36
    Van Lenthe, J. H.; Faas, S.; Snijders, J. G. Gradients in the Ab Initio Scalar Zeroth-Order Regular Approximation (ZORA) Approach. Chem. Phys. Lett. 2000, 328, 107112,  DOI: 10.1016/S0009-2614(00)00832-0
  37. 37
    Rondina, G. G.; Da Silva, J. L. F. A Revised Basin-Hopping Monte Carlo Algorithm for Structure Optimization of Clusters and Nanoparticles. J. Chem. Inf. Model. 2013, 53, 22822298,  DOI: 10.1021/ci400224z
  38. 38
    Vargas, A.; Santarossa, G.; Iannuzzi, M.; Baiker, A. Fluxionality of Gold Nanoparticles Investigated by Born-Oppenheimer Molecular Dynamics. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 80, 195421,  DOI: 10.1103/PhysRevB.80.195421
  39. 39
    Hirshfeld, F. L. Bonded-Atom Fragments for Describing Molecular Charge Densities. Theoret. Chim. Acta 1977, 44, 129138,  DOI: 10.1007/BF00549096
  40. 40
    Ceriotti, M.; Tribello, G. A.; Parrinello, M. Simplifying the Representation of Complex Free-energy Landscapes Using Sketch-map. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 1302313028,  DOI: 10.1073/pnas.1108486108
  41. 41
    Ceriotti, M.; Tribello, G. A.; Parrinello, M. Demonstrating the Transferability and the Descriptive Power of Sketch-map. J. Chem. Theory Comput. 2013, 9, 15211532,  DOI: 10.1021/ct3010563
  42. 42
    Shirts, M. R.; Chodera, J. D. Statistically Optimal Analysis of Samples from Multiple Equilibrium States. J. Chem. Phys. 2008, 129, 124105,  DOI: 10.1063/1.2978177
  43. 43
    Hoppe, R. The Coordination Number – an ”Inorganic Chameleon. Angew. Chem., Int. Ed. Engl. 1970, 9, 2534,  DOI: 10.1002/anie.197000251
  44. 44
    Da Silva, J. L. F. Effective Coordination Concept Applied for Phase Change (GeTe)m(Sb2Te3)n Compounds. J. Appl. Phys. 2011, 109, 023502  DOI: 10.1063/1.3533422

Cited By

ARTICLE SECTIONS
Jump To

This article is cited by 8 publications.

  1. Harold P. Hodgins, Jonathan J. Foley, IV, Alexander O. Govorov, Stephen K. Gray, Ian P. Hamilton. Emergence of a Low-Energy Peak in the Smoothed Absolute Value Circular Dichroism Spectra of Small Helical Gold Nanorods. ACS Omega 2024, 9 (5) , 5224-5229. https://doi.org/10.1021/acsomega.3c04943
  2. João Paulo A. de Mendonça, Felipe V. Calderan, Tuanan C. Lourenço, Marcos G. Quiles, Juarez L. F. Da Silva. Theoretical Framework Based on Molecular Dynamics and Data Mining Analyses for the Study of Potential Energy Surfaces of Finite-Size Particles. Journal of Chemical Information and Modeling 2022, 62 (22) , 5503-5512. https://doi.org/10.1021/acs.jcim.2c00957
  3. Pilar Carro, Luis Miguel Azofra, Tim Albrecht, Roberto C. Salvarezza, Evangelina Pensa. Unraveling the Causes of the Instability of Aun(SR)x Nanoclusters on Au(111). Chemistry of Materials 2021, 33 (9) , 3428-3435. https://doi.org/10.1021/acs.chemmater.1c00816
  4. Juan C. Burgos, Sol M. Mejía, Gregory F. Metha. Effect of Charge and Phosphine Ligands on the Electronic Structure of the Au8 Cluster. ACS Omega 2019, 4 (5) , 9169-9180. https://doi.org/10.1021/acsomega.9b00225
  5. Diego Guedes-Sobrinho, Renato P. Orenha, Renato L. T. Parreira, Glaucio R. Nagurniak, Gabriel Reynald Da Silva, Maurício J. Piotrowski. The effect of different energy portions on the 2D/3D stability swapping for 13-atom metal clusters. Physical Chemistry Chemical Physics 2022, 24 (11) , 6515-6524. https://doi.org/10.1039/D2CP00148A
  6. Maurício J. Piotrowski, Renato P. Orenha, Renato L. T. Parreira, Diego Guedes‐Sobrinho. Assessment of the van der Waals, Hubbard U parameter and spin‐orbit coupling corrections on the 2D / 3D structures from metal gold congeners clusters. Journal of Computational Chemistry 2022, 43 (4) , 230-243. https://doi.org/10.1002/jcc.26784
  7. Craig A. Vandervelden, Salman A. Khan, Baron Peters. Importance learning estimator for the site-averaged turnover frequency of a disordered solid catalyst. The Journal of Chemical Physics 2020, 153 (24) https://doi.org/10.1063/5.0037450
  8. Juan I. Rodríguez, Ulises A. Vergara-Beltran, Jochen Autschbach. Size evolution study on the electronic and optical properties of gold-cluster complexes Au 4 -S- C n H 2 n - S ′ - Au 4 ′ (n = 2–5). Chemical Physics Letters 2019, 732 , 136625. https://doi.org/10.1016/j.cplett.2019.136625
  • Abstract

    Figure 1

    Figure 1. Relative total energies and configurations used as starting points of the MD trajectories. Structures labeled as Au38J, Au40J1, and Au40J2 are from ref (19), while structures labeled as Au25, Au38, and Au40 are found via EAM-RBHMC and further optimized with PBE+MBD (see text for details). The core (shell) atoms are indicated by blue (yellow). The labeling Auc@Aus marks the number of atoms in the core (c) and shell (s).

    Figure 2

    Figure 2. (a) Radial distribution function, g(r), and cutoff distances (solid vertical lines) with respect to the cluster’s center of mass defining the core, shell, and protrusion regions for the Au38 nanocluster at 600 and 300 K. (b) Evolution of the Hirshfeld charges for core and shell atoms and (c) evolution of number of atoms belonging to protrusions (green line), shell (orange line), and core (blue line) regions at 600 and 300 K for Au38. Panels d–f show the same quantities as in panels a–c, respectively, but for Au40.

    Figure 3

    Figure 3. (a) Average bond lifetime, τ, and sample standard deviation (bars) among the core–core atoms (τcc), core–shell atoms (τcs), shell–shell atoms (τss), and shell-protrusions (τps) atoms, for the Au25, Au38, and Au40 nanoclusters as a function of the sampled temperatures. The lines are a guide to the eye.

    Figure 4

    Figure 4. Probability of cluster configurations at 300 K, calculated via MBAR as a function of the sketch-map coordinates for the population sampled for Au38 (left side, based on a total of 400 ps of MD sampling) and Au40 (right side, based on 500 ps MD). The selected structures distinguish core (blue), shell (yellow), and protrusion (green) regions. For each selected structure, also the corresponding 13-dimensional coordination-histogram, the input for the sketch-map dimensionality reduction, is shown. The dots mark structures selected along the MD runs, positioned in the areas of highest probability, and the stars locate the corresponding geometrically relaxed structures (dot/star of the same color indicate initial/final structures for a geometry relaxation). For Au38, the pink star is the pGMC found by us (see its structure in Figure S15). The pGMC identified in ref (19) is marked by a diamond. For Au40, the red star is the pGMC found by us, which is also the pGMC identified in ref (19).

    Figure 5

    Figure 5. Left (right): Relative stability of the tetrahedral (trigonal-bipyramidal) structure with respect to the planar Au4 (Au5) structure, as a function of the positive total charge of the cluster.

  • References

    ARTICLE SECTIONS
    Jump To

    This article references 44 other publications.

    1. 1
      III Aiken, J. D.; Finke, R. G. A. A Review of Modern Transition-Metal Nanoclusters: Their Synthesis, Characterization, and Applications in Catalysis. J. Mol. Catal. A: Chem. 1999, 145, 144,  DOI: 10.1016/S1381-1169(99)00098-9
    2. 2
      Jin, R. Atomically Precise Metal Nanoclusters: Stable Sizes and Optical Properties. Nanoscale 2015, 7, 15491565,  DOI: 10.1039/C4NR05794E
    3. 3
      Marjomäki, V.; Lahtinen, T.; Martikainen, M.; Koivisto, J.; Malola, S.; Salorinne, K.; Pettersson, M.; Häkkinen, H. Site-Specific Targeting of Enterovirus Capsid by Functionalized Monodisperse Gold Nanoclusters. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 12771281,  DOI: 10.1073/pnas.1310973111
    4. 4
      Chaves, A. S.; Piotrowski, M. J.; Da Silva, J. L. F. Structural, Energetics, and Electronic Properties Evolution of the 3d, 4d, and 5d Transition-Metal Clusters (30 TMn Systems for n= 2 – 15): A Density Functional Theory Investigation. Phys. Chem. Chem. Phys. 2017, 19, 1548415502,  DOI: 10.1039/C7CP02240A
    5. 5
      Mostafa, S.; Behafarid, F.; Croy, J. R.; Ono, L. K.; Li, L.; Yang, J. C.; Frenkel, A. I.; Cuenya, B. R. Shape-Dependent Catalytic Properties of Pt Nanoparticles. J. Am. Chem. Soc. 2010, 132, 1571415719,  DOI: 10.1021/ja106679z
    6. 6
      Chaves, A. S.; Rondina, G. G.; Piotrowski, M. J.; Tereshchuk, P.; Da Silva, J. L. F. The Role of Charge States in the Atomic Structure of Cun and Ptn (n = 2 – 14 Atoms) Clusters: A DFT Investigation. J. Phys. Chem. A 2014, 118, 1081310821,  DOI: 10.1021/jp508220h
    7. 7
      Jadzinsky, P. D.; Calero, G.; Ackerson, C. J.; Bushnell, D. A.; Kornberg, R. D. Structure of a Thiol Monolayer-Protected Gold Nanoparticle at 1.1 Å Resolution. Science 2007, 318, 430433,  DOI: 10.1126/science.1148624
    8. 8
      Yamauchi, H.; Ito, S.; Yoshida, K.-i.; Itoh, T.; Tsuboi, Y.; Kitamura, N.; Miyasaka, H. Temperature Near Gold Nanoparticles under Photoexcitation: Evaluation Using a Fluorescence Correlation Technique. J. Phys. Chem. C 2013, 117, 83888396,  DOI: 10.1021/jp311173j
    9. 9
      Sugimura, Y.; Suzuki, Y. A. A High-Temperature Catalytic Oxidation Method for the Determination of Non-Volatile Dissolved Organic Carbon in Seawater by Direct Injection of a Liquid Sample. Mar. Chem. 1988, 24, 105131,  DOI: 10.1016/0304-4203(88)90043-6
    10. 10
      Haruta, M. When Gold Is Not Noble: Catalysis by Nanoparticles. Chem. Rec. 2003, 3, 7587,  DOI: 10.1002/tcr.10053
    11. 11
      Häkkinen, H.; Barnett, R. N.; Landman, U. Electronic Structure of Passivated Au38 (SCH3)24 Nanocrystal. Phys. Rev. Lett. 1999, 82, 32643267,  DOI: 10.1103/PhysRevLett.82.3264
    12. 12
      Heaven, M. W.; Dass, A.; White, P. S.; Holt, K. M.; Murray, R. W. Crystal Structure of the Gold Nanoparticle [N(C8 H17)4][Au25 (SCH2 CH2 Ph)18]. J. Am. Chem. Soc. 2008, 130, 37543755,  DOI: 10.1021/ja800561b
    13. 13
      Farrag, M.; Tschurl, M.; Dass, A.; Heiz, U. Infra-Red Spectroscopy of Size Selected Au25, Au38 and Au144 Ligand Protected Gold Clusters. Phys. Chem. Chem. Phys. 2013, 15, 12539,  DOI: 10.1039/c3cp51406d
    14. 14
      Azubel, M.; Koivisto, J.; Malola, S.; Bushnell, D.; Hura, G. L.; Koh, A. L.; Tsunoyama, H.; Tsukuda, T.; Pettersson, M.; Häkkinen, H.; Kornberg, R. D. Electron Microscopy of Gold Nanoparticles at Atomic Resolution. Science 2014, 345, 909912,  DOI: 10.1126/science.1251959
    15. 15
      Jin, R.; Zeng, C.; Zhou, M.; Chen, Y. Atomically Precise Colloidal Metal Nanoclusters and Nanoparticles: Fundamentals and Opportunities. Chem. Rev. 2016, 116, 1034610413,  DOI: 10.1021/acs.chemrev.5b00703
    16. 16
      Garzón, I. L.; Beltrán, M. R.; González, G.; Gutierrez-González, I.; Michaelian, K.; Reyes-Nava, J. A.; Rodriguez-Hernández, J. I. Chirality, defects, and disorder in gold clusters. Eur. Phys. J. D 2003, 24, 105109,  DOI: 10.1140/epjd/e2003-00187-4
    17. 17
      Xing, X.; Yoon, B.; Landman, U.; Parks, J. H. Structural Evolution of Au Nanoclusters: From Planar to Cage to Tubular Motifs. Phys. Rev. B: Condens. Matter Mater. Phys. 2006, 74, 165423,  DOI: 10.1103/PhysRevB.74.165423
    18. 18
      Bulusu, S.; Li, X.; Wang, L.-S.; Zeng, X. C. Structural Transitions from Pyramidal to Fused Planar to Tubular to Core/Shell Compact in Gold Clusters: Aun (n= 21–25). J. Phys. Chem. C 2007, 111, 41904198,  DOI: 10.1021/jp068897v
    19. 19
      Jiang, D.; Walter, M. Au40: A Large Tetrahedral Magic Cluster. Phys. Rev. B: Condens. Matter Mater. Phys. 2011, 84, 193402,  DOI: 10.1103/PhysRevB.84.193402
    20. 20
      Wang, L.; Wang, L. Probing the Electronic Properties and Structural Evolution of Anionic Gold Clusters in the Gas Phase. Nanoscale 2012, 4, 4038,  DOI: 10.1039/c2nr30186e
    21. 21
      Zeng, C.; Chen, Y.; Das, A.; Jin, R. Transformation Chemistry of Gold Nanoclusters: From One Stable Size to Another. J. Phys. Chem. Lett. 2015, 6, 29762986,  DOI: 10.1021/acs.jpclett.5b01150
    22. 22
      Shao, N.; Huang, W.; Mei, W.-N.; Wang, L. S.; Wu, Q.; Zeng, X. C. Structural Evolution of Medium-Sized Gold Clusters Aun–(n = 36, 37, 38): Appearance of Bulk-Like Face Centered Cubic Fragment. J. Phys. Chem. C 2014, 118, 68876892,  DOI: 10.1021/jp500582t
    23. 23
      Pande, S.; Huang, W.; Shao, N.; Wang, L.; Khetrapal, N.; Mei, W.; Jian, T.; Wang, L.; Zeng, X. C. Structural Evolution of Core–Shell Gold Nanoclusters: Aun (n = 42 – 50). ACS Nano 2016, 10, 1001310022,  DOI: 10.1021/acsnano.6b04330
    24. 24
      Cleveland, C. L.; Luedtke, W. D.; Landman, U. Melting of Gold Clusters. Phys. Rev. B: Condens. Matter Mater. Phys. 1999, 60, 50655077,  DOI: 10.1103/PhysRevB.60.5065
    25. 25
      de Bas, B. S.; Ford, M. J.; Cortie, M. B. Melting in Small Gold Clusters: a Density Functional Molecular Dynamics Study. J. Phys.: Condens. Matter 2006, 18, 5574,  DOI: 10.1088/0953-8984/18/1/004
    26. 26
      Kang, J.; Wei, S.; Kim, Y. Origin of the Diverse Melting Behaviors of Intermediate-Size Nanoclusters: Theoretical Study of AlN (N = 51–58, 64). J. Am. Chem. Soc. 2010, 132, 1828718291,  DOI: 10.1021/ja107683m
    27. 27
      Beret, E. C.; Ghiringhelli, L. M.; Scheffler, M. Free Gold Clusters: Beyond the Static, Monostructure Description. Faraday Discuss. 2011, 152, 153,  DOI: 10.1039/c1fd00027f
    28. 28
      Ghiringhelli, L. M.; Gruene, P.; Lyon, J. T.; Rayner, D. M.; Meijer, G.; Fielicke, A.; Scheffler, M. Not so Loosely Bound Rare Gas Atoms: Finite-Temperature Vibrational Fingerprints of Neutral Gold-Cluster Complexes. New J. Phys. 2013, 15, 083003  DOI: 10.1088/1367-2630/15/8/083003
    29. 29
      Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 38653868,  DOI: 10.1103/PhysRevLett.77.3865
    30. 30
      Tkatchenko, A.; Ambrosetti, A.; DiStasio, R. A. Interatomic Methods for the Dispersion Energy Derived from the Adiabatic Connection Fluctuation-Dissipation Theorem. J. Chem. Phys. 2013, 138, 074106  DOI: 10.1063/1.4789814
    31. 31
      Blum, V.; Gehrke, R.; Hanke, F.; Havu, P.; Havu, V.; Ren, X.; Reuter, K.; Scheffler, M. Ab Initio Molecular Simulations with Numeric Atom-Centered Orbitals. Comput. Phys. Commun. 2009, 180, 21752196,  DOI: 10.1016/j.cpc.2009.06.022
    32. 32
      Bussi, G.; Donadio, D.; Parrinello, M. Canonical Sampling Through Velocity Rescaling. J. Chem. Phys. 2007, 126, 014101  DOI: 10.1063/1.2408420
    33. 33
      Yu, C.; Harbich, W.; Sementa, L.; Ghiringhelli, L. M.; Aprá, E.; Stener, M.; Fortunelli, A.; Brune, H. Intense Fluorescence of Au20. J. Chem. Phys. 2017, 147, 074301  DOI: 10.1063/1.4996687
    34. 34
      Häkkinen, H.; Moseler, M.; Kostko, O.; Morgner, N.; Hoffmann, M. A.; v. Issendorff, B. Symmetry and Electronic Structure of Noble-Metal Nanoparticles and the Role of Relativity. Phys. Rev. Lett. 2004, 93, 093401,  DOI: 10.1103/PhysRevLett.93.093401
    35. 35
      Huang, W.; Ji, M.; Dong, C.-D.; Gu, X.; Wang, L.-M.; Gong, X. G.; Wang, L. Relativistic Effects and the Unique Low-Symmetry Structures of Gold Nanoclusters. ACS Nano 2008, 2, 897904,  DOI: 10.1021/nn800074b
    36. 36
      Van Lenthe, J. H.; Faas, S.; Snijders, J. G. Gradients in the Ab Initio Scalar Zeroth-Order Regular Approximation (ZORA) Approach. Chem. Phys. Lett. 2000, 328, 107112,  DOI: 10.1016/S0009-2614(00)00832-0
    37. 37
      Rondina, G. G.; Da Silva, J. L. F. A Revised Basin-Hopping Monte Carlo Algorithm for Structure Optimization of Clusters and Nanoparticles. J. Chem. Inf. Model. 2013, 53, 22822298,  DOI: 10.1021/ci400224z
    38. 38
      Vargas, A.; Santarossa, G.; Iannuzzi, M.; Baiker, A. Fluxionality of Gold Nanoparticles Investigated by Born-Oppenheimer Molecular Dynamics. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 80, 195421,  DOI: 10.1103/PhysRevB.80.195421
    39. 39
      Hirshfeld, F. L. Bonded-Atom Fragments for Describing Molecular Charge Densities. Theoret. Chim. Acta 1977, 44, 129138,  DOI: 10.1007/BF00549096
    40. 40
      Ceriotti, M.; Tribello, G. A.; Parrinello, M. Simplifying the Representation of Complex Free-energy Landscapes Using Sketch-map. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 1302313028,  DOI: 10.1073/pnas.1108486108
    41. 41
      Ceriotti, M.; Tribello, G. A.; Parrinello, M. Demonstrating the Transferability and the Descriptive Power of Sketch-map. J. Chem. Theory Comput. 2013, 9, 15211532,  DOI: 10.1021/ct3010563
    42. 42
      Shirts, M. R.; Chodera, J. D. Statistically Optimal Analysis of Samples from Multiple Equilibrium States. J. Chem. Phys. 2008, 129, 124105,  DOI: 10.1063/1.2978177
    43. 43
      Hoppe, R. The Coordination Number – an ”Inorganic Chameleon. Angew. Chem., Int. Ed. Engl. 1970, 9, 2534,  DOI: 10.1002/anie.197000251
    44. 44
      Da Silva, J. L. F. Effective Coordination Concept Applied for Phase Change (GeTe)m(Sb2Te3)n Compounds. J. Appl. Phys. 2011, 109, 023502  DOI: 10.1063/1.3533422
  • Supporting Information

    Supporting Information

    ARTICLE SECTIONS
    Jump To

    The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.8b03397.

    • Detailed computational details for the EAM-RBHMC, DFT-MD, and sketch-map-training calculations; the g(r) and Hirshfeld analyses for all MD trajectories; the resulting sketch-map for Au25; MBAR-estimated error bars on the probabilities reported in all the sketch-maps; detailed bond-lifetime analysis for each trajectory; detailed information on all found local minima; and further information on stability analysis of charged Au4 and Au5 (PDF)


    Terms & Conditions

    Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

Pair your accounts.

Export articles to Mendeley

Get article recommendations from ACS based on references in your Mendeley library.

Pair your accounts.

Export articles to Mendeley

Get article recommendations from ACS based on references in your Mendeley library.

You’ve supercharged your research process with ACS and Mendeley!

STEP 1:
Click to create an ACS ID

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

MENDELEY PAIRING EXPIRED
Your Mendeley pairing has expired. Please reconnect