Search

The paper reports the results of a molecular dynamics study of the heating and melting process of nanoparticles with 1985 to 84 703 atoms. Building on a previous study by the present authors [Bertoldi, et al., J. Phys. Chem. Solids, 2017, 111, 286-293] involving the energy versus temperature, the Lindemann index and the radial distribution function, the current work relies on the mean-square displacement, the Lindemann ratio and the simulated snapshots to characterize four regions in the process of heating-to-melting. A general pattern of the atomic configuration evolution upon heating and a systematics of the transition temperatures between the various identified steps, is proposed. In addition, the most significant, so-called "melting step" in this process is analyzed in terms of the quasi-chemical approach proposed by Bertoldi et al., which treats this step by invoking a dynamic equilibrium of the type Au (LEA-SPL) ⇌ Au (HEA-LPL) involving low-energy atoms (LEA) and high-energy atoms (HEA) forming the solid phase-like (SPL) and the liquid phase-like (LPL) states of the system, respectively. The "melting step" is characterized by evaluating the equal-Gibbs energy temperature, i.e., the "T0 temperature", previously introduced by the current authors, which is the thermodynamic counterpart of the temperature of fusion of macroscopic elemental solids. The diffusion coefficients at T0 are determined, and their spatial and temperature dependence is discussed. In particular, the activation energy for the atom movements in the HEA-LPL/LEA-SPL mixture at T0 is reported. The consistency between the current phenomenological picture and microscopic interpretation of the thermodynamic, kinetic and atomic configuration information obtained is highlighted.

Available macroscopic theories-such as the Johnson-Kendall-Roberts (JKR) model-predict spherical particles to stick to each other at small collision velocities v; above the bouncing velocity, v_{b}, they bounce. We study the details of the bouncing threshold using molecular dynamics simulation for crystalline nanoparticles where atoms interact via the Lennard-Jones potential. We show that the bouncing velocity strongly depends on the nanoparticle orientation during collision; for some orientations, nanoparticles stick at all velocities. The dependence of bouncing on orientation is caused by energy dissipation during dislocation activity. The bouncing velocity decreases with increasing nanoparticle radius in reasonable agreement with JKR theory. For orientations for which bouncing exists, nanoparticles stick again at a higher velocity, the fusion velocity, v_{f}, such that bouncing only occurs in a finite range of velocities-the bouncing window. The fusion velocity is rather independent of the nanoparticle radius.

Using molecular dynamics simulations, we study collisions between amorphous silica nanoparticles. Our silica model contains uncontaminated surfaces, that is, the effect of surface hydroxylation or of adsorbed water layers is excluded. For central collisions, we characterize the boundary between sticking and bouncing collisions as a function of impact velocity and particle size and quantify the coefficient of restitution. We show that the traditional Johnson-Kendall-Roberts (JKR) model provides a valid description of the ingoing trajectory of two grains up to the moment of maximum compression. The distance of closest approach is slightly underestimated by the JKR model, due to the appearance of plasticity in the grains, which shows up in the form of localized shear transformation zones. The JKR model strongly underestimates the contact radius and the collision duration during the outgoing trajectory, evidencing that the breaking of covalent bonds during grain separation is not well described by this model. The adhesive neck formed between the two grains finally collapses while creating narrow filaments joining the grains, which eventually tear.

While at small collision velocities collisions of nanoparticles (NPs) are elastic, they become plastic at higher velocities. We study the elastic-plastic threshold and the onset of plasticity using molecular dynamics simulation for a Lennard-Jones material. The reasons behind the R^{-2/3} increase of the threshold velocity for small NP radii R found recently are discussed. At the threshold, NP orientation strongly influences the generation of plasticity, and averaging over many orientations is required to predict the critical velocity for dislocation generation. The onset of plasticity is governed by the generation of isolated stacking faults and nanotwins spanning the entire NP. At higher velocities, the fraction of defects becomes proportional to the total number of atoms in the NP.

When nanoparticles (NPs) collide with low velocities, they interact elastically in the sense that--besides their fusion caused by their mutual van-der-Waals attraction--no defects are generated. We investigate the minimum velocity, vc, necessary for generating defects and inducing plasticity in the NP. The determination of this elastic-plastic threshold is of prime importance for modeling the behavior of granular matter. Using the generic Lennard-Jones interaction potential, we find vc to increase strongly with decreasing radius. Current models do not agree with our simulations, but we provide a model based on dislocation emission in the contact zone that quantitatively describes the size dependence of the elastic-plastic transition.