Multiscale modeling of fragmentation in an electrospray plume

Abstract

We present a 3D-particle-in-cell (PIC) approach to modeling electrospray plumes typical of those formed by externally wetted emitter devices. Numerical grid-resolution techniques suitable for capturing strong electric fields in the emitter region were explored, and grid refinement criteria were quantified. The molecular dynamics simulations of the EMIM−BF4 ionic liquid system were modeled to determine the fragmentation mechanism in the presence of an electric field and dimer temperature as well as to provide fragmentation rates for the PIC simulations. An energy analysis of the molecular dynamics (MD) fragmentation demonstrated that the key mechanism for dimer fragmentation corresponds to a decrease in the Coulomb energy between the cation and anion in the system and that dimers of temperatures 300 and 600 K are extremely stable for electric fields smaller than 1.5 V/nm. Using probabilities of fragmentation consistent with the MD simulations, we implemented a dimer fragmentation model in our PIC simulations. The ion energy distribution functions obtained from the PIC simulations were used to predict retarding potential analysis (RPA) curves that were compared directly to measurements. The sensitivity of the RPA shape to the fragmentation probability was found to be significant. By comparing predicted and measured RPA curves for both negative and positive operation modes, and the fact that dimers do not fragment for electric fields less than 0.6 V/nm, we conclude that fragmentation of dimers occurs spontaneously due to their high thermal energies.

Multiscale modeling of electrospray ion emission

“…The solution of the Navier-Stokes equations is solved after the electric problem [Eqs. (10) and (11)] subject to the outlet condition in Eq. ( 12) together with mixed Dirichlet and Neumann boundary conditions at the interface.…”

“…Fragmentation occurs both due to the high internal energy of ionic liquid clusters and the force of the electric field pulling the clusters apart. 10,13 Previous work by Miller and Lozano indicates that the fragmentation process for a population of clusters is an exponential decay process with an electric field-dependent constant rate that occurs on the picosecond to microsecond timescale in a typical electrospray thruster. 13,31 In the n-body model, ionic liquid cluster fragmentation is modeled as discrete events occurring each time step in the simulation.…”

“…One way to determine these initial conditions is to compare simulation outputs with experimental energy distribution data. 10,33 The energy distribution of the beam is analyzed using retarding potential analysis (RPA) curves. 13 The shape of an RPA curve is determined almost entirely by the percentage of each species in the beam and the fragmentation rates of each species as a function of electric field strength.…”

“…Thus, it is appropriate to compare the 20 ns simulated RPA curve with experimental data taken over a much longer timescale as has been done in recent work. 10 RPA curves are generated by calculating the kinetic energy of each ion cluster when they reach a plane that is designated as the detector. The distance to the detector plane is 8.85 cm, which is the same as the distance to the detector for the experimental curves used for comparison.…”

“…The integration of an electrohydrodynamic source model with a discrete plume model presents new opportunities for validation of the meniscus solution. Furthermore, it offers promise to bridge the gap between the macroscale design parameter space (emitter geometry, liquid properties) and the plume evolution via realistic initial conditions in ways that cannot be captured through molecular dynamics or simulations alone, 9 limited to domain size, or constant axial flux, 10 which are usually constrained to a limited set of operating conditions.…”