Combined molecular dynamics–direct simulation Monte Carlo computational study of laser ablation plume evolution

Abstract

A two-stage computational model of evolution of a plume generated by laser ablation of an organic solid is proposed and developed. The first stage of the laser ablation, which involves laser coupling to the target and ejection of molecules and clusters, is described by the molecular dynamics (MD) method. The second stage of a long-term expansion of the ejected plume is modeled by the direct simulation Monte Carlo (DSMC) method. The presence of clusters, which comprise a major part of the overall plume at laser fluences above the ablation threshold, presents the main computational challenge in the development of the combined model. An extremely low proportion of large-sized clusters hinders both the statistical estimation of their characteristics from the results of the MD model and the following representation of each cluster size as a separate species, as required in the conventional DSMC. A number of analytical models are proposed and verified for the statistical distributions of translational and internal energies of monomers and clusters as well as for the distribution of the cluster sizes, required for the information transfer from the MD to the DSMC parts of the model. The developed model is applied to simulate the expansion of the ablation plume ejected in the stress-confinement irradiation regime. The presence of the directly ejected clusters drastically changes the evolution of the plume as compared to the desorption regime. A one-dimensional self-similar flow in the direction normal to the ablated surface is developed within the entire plume at the MD stage. A self-similar two-dimensional flow of monomers forms in the major part of the plume by about 40 ns, while its counterpart for large clusters forms much later, leading to the plume sharpening effect. The expansion of the entire plume becomes self-similar by about 500 ns, when interparticle interactions vanish. The velocity distribution of particles cannot be characterized by a single translational temperature; rather, it is characterized by a spatially and direction dependent statistical scatter about the flow velocity. The cluster size dependence of the internal temperature is mainly defined by the size dependence of the unimolecular dissociation energy of a cluster.

Computer Simulations of Laser Ablation of Molecular Substrates

“…[131][132][133][134][135] The first simulations performed with the combined MD-DSMC approach have demonstrated the ability of the method to provide insights into the complex processes occurring during the evolution of the ablation plume. 86,87 Some of the results obtained to date are presented in Section VI.…”

“…40,42 The long-term plume expansion can be simulated by the DSMC method, part D in Figure 1, whereas the initial conditions for DSMC simulation can be provided by the MD breathing sphere model simulations. 74,77,78,86,87 A fine grid of small black points in part D of Figure 1 schematically represents the spatial resolution in the DSMC model, as discussed in Section II.E. The kinetics of chemical reactions, cluster evaporation/ growth, and ionization can be reproduced by solving a system of rate equations (shown schematically by green squares in part D of Figure 1).…”

“…74,78 One possible solution of this problem is to divide clusters into groups. 74,86,87 The range of cluster sizes that form a group can be chosen so that clusters in a group have similar velocity and spatial distributions in the plume. The characteristics that can provide a connection between MD and DSMC simulations in the multiscale model are summarized in Figure 4b.…”

“…87,175 The driving forces and characteristic times of the formation of the self-similar flow, however, are different for the axial and radial directions and even the mechanisms of the self-similar flow formation in the radial direction are different for different plume components.…”

“…The process of formation of a self-similar flow of monomers in the radial direction is governed by the pressure gradient and the formation time is on the order of several tens of nanoseconds. 87 For the large clusters, the main driving force of the radial flow formation is collisions with monomers and light clusters. Because of the rapid plume expansion in the axial direction, the collision rate quickly subsides and the largest clusters acquire low radial velocities relative to their light counterparts, 87 which results in a sharpening of the overall density distribution of the ejected plume.…”