Structural classification of Ag and Cu nanocrystals with machine learning
We use machine learning (ML) to classify the structures of mono-metallic Cu and Ag nanoparticles.
Engineering topics
Publications and source records attributed to Fichthorn, Kristen A..
We use machine learning (ML) to classify the structures of mono-metallic Cu and Ag nanoparticles.
Parallel-tempering molecular dynamics simulations reveal the temperature- and size-dependent shapes of Cu nanocrystals.
Not Available
Twin structures possess distinct physical and chemical properties by virtue of their specific twin configuration. However, twinning and detwinning processes are not fully understood on the atomic scale. Integrating in situ high resolution transmission electron microscopy and molecular dynamic simulations, here, we find tensile strain in the asymmetrical 5-fold twins of Au nanoparticles leads to twin boundary migration through dislocation sliding (slipping of an atomic layer) along twin boundaries and dislocation reactions at the 5-fold axis under an electron beam. Migration of one or two layers of twin planes is governed by energy barriers, but overall, the total energy, including surface, lattice strain, and twin boundary energy, is relaxed after consecutive twin boundary migration, leading to a detwinning process. In addition, surface rearrangement of 5-fold twinned nanoparticles can aid in the detwinning process.
Here, we use two variants of replica-exchange molecular dynamics (MD) simulations, parallel tempering MD and partial replica exchange MD, to probe the minimum free-energy shapes of Ag nanocrystals containing 100–200 atoms in a vacuum, ethylene glycol (EG) solvent, and EG solvent with a PVP polymer containing 100 repeat units. Our simulations reveal a shape intermediate between a Dh and an Ih, a Dh-Ih, that has distinct structural signatures and magic sizes. We find several prominent features associated with entropy: pure FCC nanocrystals are less common than FCC crystals containing stacking faults, and crystals with the minimum potential energy are not always preferred over the range of relevant temperatures. The shapes of the nanocrystals in solution are influenced by the chemical identities of the solution-phase molecules. Comparing Ag nanocrystal shapes in EG to those in an EG+PVP solution, we find more icosahedra in EG and more decahedra in EG+PVP across all of the nanocrystal sizes probed in this study. At certain critical sizes, nanocrystal shapes can change dramatically with the addition and removal of a single atom or with a change in temperature at a fixed size. The information in our study could be useful in efforts to devise processing routes to achieve selective nanocrystal shapes.
Crystals with penta-twinned structures can be produced from diverse fcc metals, but the mechanisms that control the final product shapes are still not well understood. By using the theory of absorbing Markov chains to account for the growth of penta-twinned decahedral seeds via atom deposition and surface diffusion, we predicted the formation of various types of products: decahedra, nanorods, and nanowires. We showed that the type of product depends on the morphology of the seed and that small differences between various seed morphologies can lead to significantly different products. For the case of uncapped decahedra seeds, we compared predictions from our model to nanowire morphologies obtained in two different experiments and obtained favorable agreement. Finally, possible extensions of our model are indicated.
Penta-twinned metal nanowires are finding widespread application in existing and emerging technologies. However, little is known about their growth mechanisms. Here, we probe the origins of chloride- and alkylamine-mediated, solution-phase growth of penta-twinned Cu nanowires from first principles using multi-scale theory. Using quantum density functional theory (DFT) calculations, we characterize the binding and surface diffusion of Cu atoms on chlorine-covered Cu(100) and Cu(111) surfaces. We find stronger binding and slower diffusion of Cu atoms on chlorinated Cu(111) than on chlorinated Cu(100) – a reversal of the trend for bare Cu surfaces. We also probe inter-facet diffusion and find this proceeds faster from Cu(100) to Cu(111) than the reverse. Using the DFT rates for hopping between individual sites at Ångstrom scales, we calculate coarse-grained, inter-facet rates for nanowires of various lengths – up to hundreds of μm – and diameters in the 10-nm range. We predict nanowires with aspect ratios around 100 based on surface diffusion, alone. We also account for the influence of a self-assembled alkylamine layer that covers most of the {100} facets, but is absent or thin and disordered on the {111} facets and in an “end zone” near the {100}-{111} boundary. With an end zone, we predict a wide range of nanowire aspect ratios in the experimental ranges. Our work reveals the mechanisms by which a halide – chloride – promotes the growth of high-aspect-ratio nanowires.
Oriented attachment is a nonclassical crystal growth mechanism that frequently leads to the formation of low-dimensional nanocrystals, such as nanowires or nanoplates, as well as intricately branched nanocrystals. This chapter discusses several experimentally observed instances of such structures, as well as theoretical efforts to understand them. Recent molecular-dynamics simulation studies in our group aimed at understanding how oriented attachment could promote the aggregation of Ag nanoplates to produce two-dimensional sheets are highlighted. Here, the mechanism found in this study may extend to other systems for which lowdimensional crystal growth occurs.
The shape-controlled synthesis of Cu nanocrystals can benefit a wide range of applications, though challenges exist in achieving high and selective yields to a particular shape. Capping agents play a pivotal role in controlling shape, but their exact role remains ambiguous. In this study, the adsorption of ethylenediamine (EDA) on Cu(100) and Cu(111) was investigated with quantum density functional theory (DFT) to reveal the complex roles of EDA in promoting penta-twinned Cu nanowire growth. We find EDA has stronger binding on Cu(100) than on Cu(111), which agrees the general expectation that penta-twinned Cu nanowires express facets with stronger capping-molecule binding. Despite this stronger binding, ab initio thermodynamics reveals the surface energy of EDA-covered Cu(111) is lower than that EDA-covered Cu(100) at all solution-phase EDA chemical potentials, so there is no thermodynamic driving force for penta-twinned nanowires. We also investigated the capability of EDA to protect Cu surfaces from oxidation in water by quantifying energy barriers for a water molecule to diffuse through EDA layers on Cu(100) and Cu(111). The energy barrier on Cu(100) is significantly lower, which supports observations of faster oxidation of Cu(100) in electrochemical experiments. Thus, we elucidate another possible function of a capping agent – to enable selective oxidation of crystal facets. This finding adds to the general understanding of successful attributes of capping agents for shape-selective nanocrystal growth.
Copper nanocrystals are often grown with the help of alkylamine capping agents, which direct the nanocrystal shape. However, the role of these molecules is still unclear. We characterized the assembly of aqueous tetradecylamine (TDA) around a Cu nanocrystal and found that TDA exhibits a temperature-dependent bilayer structure. The bilayer involves an inner layer, in which TDA binds to Cu via the amine group and tends to orient the alkyl tail perpendicular to the surface, and an outer layer whose structure depends on temperature. At low temperatures, alkylamines in the inner layer form bundles with no apparent relation to the crystal facets. Alkylamines in the outer layer tend to orient their long axes perpendicular to the Cu surfaces, with interdigitation into the inner layer. At high temperatures, alkylamines in the inner layer lose their bundle structure, and outer-layer alkylamines tend to orient themselves tangential to the Cu surfaces, forming a “web” above inner-layer TDA. TDA exhibits a rapid interlayer exchange at typical synthesis temperatures, consistent with experiment. Here, the variety in the assemblies seen here and in other studies of alkanethiols around gold nanocrystals indicates a richness in the assemblies that can be achieved by modulating the interaction between the strongly binding end group and the surface.