No abstract

Atomic level simulations supported by density-functional theory calculations identify the key mechanisms of the twinning process in gold tetrahedral nanoparticles, which is shown to originate from the growth kinetics of the pure, ligand-free metal.

The lowest-energy structures of AgCu nanoalloys are searched for by global optimization algorithms for sizes 100 and 200 atoms depending on composition, and their structures and mixing energy are analyzed by machine learning tools.

The shape of AuPd nanoparticles is engineered by surface stress relaxation, achieved by varying the Au content in nanoparticles of Pd-rich compositions. AuPd nanoparticles are grown in the gas phase for several compositions and growth conditions. Their structure is atomically resolved by HRTEM/STEM and EDX. In pure Pd distributions the dominant structures are FCC truncated octahedra (TO), while increasing the Au content there is a transition to icosahedral (Ih) structures in which Au atoms are preferentially placed at the nanoparticle surface. The transition is sharper for growth conditions closer to equilibrium. The physical origin of the transition is determined with the aid of computer simulations. Global optimization searches and free energy calculations confirm that Ih become the equilibrium structure for increasing the Au content. Atomic stress calculations demonstrate that the TO → Ih shape change is caused by a better relaxation of anisotropic surface stress in icosahedra.

The gas-phase growth of AuAg and PtPd clusters up to sizes $\sim$3 nm is simulated by Molecular Dynamics. Both systems are characterized by a very small size mismatch and by a tendency of the less cohesive element to segregate at the nanoparticle surface. The aim of this work is to figure out the differences in the behavior between these two bimetallic systems at the atomic level. For each system, three simulation types are performed, in which either one species or both species are deposited on preformed bimetallic seeds. Our results show that core@shell and intermixed chemical ordering arrangements can be obtained, in agreement with the available experimental data. In the case of core@shell arrangement, the purity of the surface layer is perfect for Ag-rich and Pd-rich nanoparticles, whereas in Au-rich and Pt-rich ones, some tendency to surface migration of minority atoms (Ag or Pd) is observed. This tendency is somewhat stronger for Ag than for Pd. The analysis of the internal arrangement of the nanoparticles indicates that in the growth process the mobility of Pd and Ag minority atoms is stronger than that of Au and Pt minority atoms.

We propose a scheme for the automatic separation (i.e., clustering) of data sets composed of several nanoparticle (NP) structures by means of Machine Learning techniques. These data sets originate from atomistic simulations, such as global optimizations searches and molecular dynamics simulations, which can produce large outputs that are often difficult to inspect by hand. By combining a description of NPs based on their local atomic environment with unsupervised learning algorithms, such as K-Means and Gaussian mixture model, we are able to distinguish between different structural motifs (e.g., icosahedra, decahedra, polyicosahedra, fcc fragments, twins, and so on). We show that this method is able to improve over the results obtained previously thanks to the successful implementation of a more detailed description of NPs, especially for systems showing a large variety of structures, including disordered ones.

PtPd nanoparticles are among the most widely studied nanoscale systems, mainly because of their applications as catalysts in chemical reactions. In this work, a combined experimental-theoretical study is presented about the dependence of growth shape of PtPd alloy nanocrystals on their composition. The particles are grown in the gas phase and characterized by STEM-HRTEM. PtPd nanoalloys present a bimodal size distribution. The size of the larger population can be tuned between 3.8 ± 0.4 and 14.1 ± 2.0 nm by controlling the deposition parameters. A strong dependence of the particle shape on the composition is found: Pd-rich nanocrystals present more rounded shapes whereas Pt-rich ones exhibit sharp tips. Molecular dynamics simulations and excess energy calculations show that the growth structures are out of equilibrium. The growth simulations are able to follow the growth shape evolution and growth pathways at the atomic level, reproducing the structures in good agreement with the experimental results. Finally the optical absorption properties are calculated for PtPd nanoalloys of the same shapes and sizes grown in our experiments.

In this short communication we describe the results obtained from the application of the Gaussian Mixture Model, a popular unsupervised learning algorithm, to some modified data sets gained after the global optimizations of three different AgCu nanoalloys. In particular we highlight both positive and negative aspects of such an approach to this kind of data. We show indeed that thanks to the Common Neighbor Analysis we are still able to describe nanoalloys well enough to exploit a physically meaningful separation in different structural families, even with a very low-dimensional representation. On the other hand, we show that the imposition of an energy cutoff over the data set is a delicate matter since it forces us to find a tradeoff between having a large set of data and having clean data.