Hydrodynamic model for picosecond propagation of laser-created nanoplasmas

Abstract

a b s t r a c tThe interaction of a free-electron-laser pulse with a moderate or large size cluster is known to create a quasi-neutral nanoplasma, which then expands on hydrodynamic timescale, i.e., > 1 ps. To have a better understanding of ion and electron data from experiments derived from laser-irradiated clusters, one needs to simulate cluster dynamics on such long timescales for which the molecular dynamics approach becomes inefficient. We therefore propose a two-step Molecular Dynamics-Hydrodynamic scheme. In the first step we use molecular dynamics code to follow the dynamics of an irradiated cluster until all the photo-excitation and corresponding relaxation processes are finished and a nanoplasma, consisting of ground-state ions and thermalized electrons, is formed. In the second step we perform long-timescale propagation of this nanoplasma with a computationally efficient hydrodynamic approach.In the present paper we examine the feasibility of a hydrodynamic two-fluid approach to follow the expansion of spherically symmetric nanoplasma, without accounting for the impact ionization and three-body recombination processes at this stage. We compare our results with the corresponding molecular dynamics simulations. We show that all relevant information about the nanoplasma propagation can be extracted from hydrodynamic simulations at a significantly lower computational cost when compared to a molecular dynamics approach. Finally, we comment on the accuracy and limitations of our present model and discuss possible future developments of the two-step strategy.

Hydrodynamic model for expansion and collisional relaxation of x-ray laser-excited multi-component nanoplasma

“…Timescales of collisional processes, their interplay in time, and, finally, the expansion dynamics agree qualitatively with observations from previous simulations and experiments. 4,6,11,18 To emphasize, the hydrodynamics simulations were performed on a single CPU for a large Ar cluster (>80000 atoms) and took around 24 h for 1 ps simulation. This is significantly faster than the previously quoted MD calculations which-to remind-took 48 h on a single CPU for a 1.5 ps long simulation of Ar 1000 cluster.…”

“…The set of equations presented in Sec. II is an extended version of the model presented in [11]. The inclusion of the new source terms representing the impact ionization and three body recombination of these collisional processes in our previous model does not change the basic properties of these equations.…”

“…The inclusion of the new source terms representing the impact ionization and three body recombination of these collisional processes in our previous model does not change the basic properties of these equations. We can therefore use the same numerical methods as in [11] to solve the new set of equations: (i) the flux corrected transport (FCT) scheme [23,24] for the 0th and 1st moment equations, (ii) 4 th -order Runge-Kutta integration [25] for the fluid temperature equations, and (iii) tri-diagonal method [25] for the Poisson's equation. Moreover, to ensure the stability against sharp gradients, we use the artificial viscosity concept introduced by Lapidus [26,27].…”

“…6 In contrast, the hydrodynamic approach (HYDRO) can achieve much better computational efficiency at simulating long-timescale evolution of such systems. Saxena et al 11 have shown a comparison of the two approaches, the MD one and the hydrodynamic one, for a model system involving >10 5 particles. However, the hydrodynamic model used there enabled only sample propagation and neglected collisional relaxation processes.…”

“…In the present work we extend the model described in [11] to include impact ionization and three-body recombination processes. The extended scheme can treat a multi-component nanoplasma, for which electrons, neutrals (atoms) and ground state ions are modeled as separate fluids intermixed with each other.…”