Influence of string-like cooperative atomic motion on surface diffusion in the (110) interfacial region of crystalline Ni

Abstract

Although we often think about crystalline materials in terms of highly organized arrays of atoms, molecules, or even colloidal particles, many of the important properties of this diverse class of materials relating to their catalytic behavior, thermodynamic stability, and mechanical properties derive from the dynamics and thermodynamics of their interfacial regions, which we find they have a dynamics more like glass-forming (GF) liquids than crystals at elevated temperatures. This is a general problem arising in any attempt to model the properties of naturally occurring crystalline materials since many aspects of the dynamics of glass-forming liquids remain mysterious. We examine the nature of this phenomenon in the "simple" case of the (110) interface of crystalline Ni, based on a standard embedded-atom model potential, and we then quantify the collective dynamics in this interfacial region using newly developed methods for characterizing the cooperative dynamics of glass-forming liquids. As in our former studies of the interfacial dynamics of grain-boundaries and the interfacial dynamics of crystalline Ni nanoparticles (NPs), we find that the interface of bulk crystalline Ni exhibits all the characteristics of glass-forming materials, even at temperatures well below the equilibrium crystal melting temperature, T m . This perspective offers a new approach to modeling and engineering the properties of crystalline materials. C 2015 AIP Publishing LLC. [http://dx

Superionic UO2: A model anharmonic crystalline material

“…As we have observed before in the case of Ni and H2O interfacial regions, heating UO2 of simulated crystal with a crystal-vacuum interface leads to the formation of a mobile interfacial layer in which the Debye-Waller factor is enhanced relative to the bulk material modeled with periodic boundary conditions and a greatly accelerated diffusion for temperatures above the Tammann temperature, which in the present instance is close to T  . We illustrate the diffusion coefficient in the mobile interfacial region, defined as before 126,127 , in Figure 19, where we see a We note that the ratio of Ds int (O) to Ds(O), quantifying the interfacial acceleration of diffusion in this interfacial region of UO2, is much smaller than observed in previous simulations of heated crystals. Tentatively, we associate this trend with the high cohesive energy of ionic versus non-ionic materials 83 (see discussion above), which can be expected to diminish the relative amplitude of atomic motion in the interfacial region and lead to a general tendency of ionic crystalline materials to superheat.…”

“…[122][123][124][125] In general, these changes in mobility are highly grain boundary specific and previous work on the grain boundaries of different types of Ni 106 has shown that this variability reflects the variable extent of collective motion in these different types of interfacial regions and a similar highly variable mobility has been observed in the interfacial regions defined by the different crystallographic interfaces of an isolated Ni crystal. 126 Given the importance of interfacial mobility of superionic materials in applications, we briefly consider the interfacial mobility on the (110) interface of UO2.…”

String-like collective motion in the α- and β-relaxation of a coarse-grained polymer melt

“…Cicerone and coworkers have suggested that the Johari-Goldstein relaxation process corresponds to transition events between metabasins in the potential energy landscape 91 and further that the rate of these transitions governs the rate of molecular diffusion. 92 This proposed interpretation of the Johari-Goldstein relaxation process, and the finding that t ⇤ scales inversely to the molecular diffusion coefficient in small molecule fluids, 37,[49][50][51][52] would suggest that t ⇤ might be identified with the Johari-Goldstein or slow -relaxation process. This is a curious possibility that would naturally explain the observed decoupling relation between ⌧ ↵ and t ⇤ found in previous simulation studies.…”

“…1,2 molecule liquids, the time scale t ⇤ has been observed to scale inversely with the diffusion coefficient so that this time can be interpreted as "diffusive time scale." 37,[49][50][51][52] Of course, this interpretation is more complicated in polymer fluids because of chain connectivity, and in such fluids, t ⇤ is a relaxation time associated with segmental displacement motion. The characteristic time t ⇤ is generally shorter than the ↵-relaxation time, ⌧ ↵ , but longer than ⌧ .…”

Comparative Study of the Collective Dynamics of Proteins and Inorganic Nanoparticles

“…The small size of NPs and their relatively large surface area normally imply a significant downward shift of the melting temperature of the NPs 5 6 7 8 9 10 11 12 and appreciable excitation of motion on the surfaces of these particles at T well below the melting point temperature, T m , of the bulk material 4 . Even at T as much as 30% to 50% below the NP T m , these nano-crystals exhibit a relatively high interfacial mobility, although it is simplistic to characterize their interfacial dynamics as being like a ‘liquid’ 13 14 15 16 . This ‘pre-melted’ 5 12 17 condition makes the interfacial regions of metal NPs highly mobile 13 , and significant fluctuations in particle shape can emerge when the NPs actually have a size comparable to 1 nm 18 .…”