Session GT2: Electron-Molecule and Plasma-Related Collisions

GEC Early Career Award: Development of Collision Models and Data Through to Applications in Plasma Modeling

Modeling low-temperature non-equilibrium plasmas involves complex physics and computational algorithms that require the coupling of particle kinetics (including collisional and

radiative processes) and electric and magnetic fields. Over the last decade or so, collisional-radiative (CR), Monte Carlo (MC) particle-in-cell (MCPIC) and Boltzmann solver codes have

become the primary tools of use to model these plasmas. However, these tools rely on input kinetic data and models: cross sections, radiative-data and collision models, which are

particularly difficult to calculate for low-temperature plasmas, where near-neutral atoms and molecules, and excited state species are abundant. As a result, in general, comprehensive sets of kinetic data and models do not exist in the literature. In this talk we review the steps taken to generate kinetic data and analyze their impact in

plasma modeling tools. Specifically, we will review a recently developed frame-work for quickly calculating radiative and electron-impact cross sections that can been applied to near-neutral ground and excited states species for atoms, ions and diatomic molecules (to be applied to polyatomics in the future). We compare these data to fundamental cross section experiments, and utilizing our recently developed MC code “ThunderBoltz” perform swarm-type calculations to validate the data sets.   

Studies of chiral electron scattering by chiral molecules using a Rb spin exchange source

Using a Rb spin filter as a polarized electron source [1], we performed experiments searching for a chirality-dependent secondary electron yield when a 141 eV longitudinally spin-polarized electron beam was incident on a solid cysteine target with randomly oriented molecules. We determined the secondary electron yield by measuring the positive current produced when the cysteine target was negatively biased. We found no spin dependent effects at a level of one part in one thousand from this interaction, a collision channel that has not been studied to date. This experiment will be discussed in the context of our understanding of the electromagnetic dynamics responsible for chiral dependence in electron-molecule collisions.

[1] M. Pirbhai, J. Knepper, E. T. Litaker, D. Tupa, and T. J. Gay, “Optically pumped spinexchange polarized-electron source,” Phys.Rev. A 88R, 1 (2013).

    

A Machine Learning Model for Predicting Electron-Molecule Ionization Cross Sections

Electron-molecule collision cross sections play a pivotal role in many areas of applied physics, including plasma physics.  Unfortunately, detailed measurements and state-of-the-art theoretical calculations are often too difficult to perform for highly complex molecules or for a wide variety of molecular targets.  As a result, the data sets needed for modeling applications are often unavailable or incomplete.  Machine learning algorithms may be able to help fill the gap in available cross section data and provide reasonable estimates for molecular targets that are beyond the reach of theoretical models and experimental measurements.  We present a feed-forward neural network trained on existing experimental data and show that it provides reasonable estimates of electron-molecule collision cross sections for molecular targets beyond those in its training set.  Our model is benchmarked by comparing its predictions with measured cross sections, and our data demonstrate that with training on as a few as 15 molecular targets, the algorithm predicts the measured cross sections to within 10% in many cases.  The success of our simple model with relatively few training data sets indicates that machine learning techniques can successfully complement traditional theoretical models and experiment measurements and represent a viable method to provide much needed data for plasma modeling.   

Ionization of molecules using a Gaussian representation of the continuum states

We are interested in investigating single ionization of small molecules by photon or electron impact. The focus of our work is on the description of the electron ejected into the continuum.

The theoretical study of molecular ionization processes requires accurate multicentric wavefunctions. Because of their very interesting mathematical properties, Gaussian basis sets are widely exploited for bound states calculations, while their use is far less developed in the context of fragmentation processes where continuum states play a leading role. Nodeless real Gaussian-type orbitals (rGTO) are not suited to represent oscillating and non-decreasing continuum wavefunctions. On the other hand, complex GTO (cGTO) --i.e. Gaussians with a complex exponent-- intrinsically oscillate and should be more adapted for this task.

Recently [1], we have explored this lead and developed an efficient non-linear optimization method to provide sets of cGTOs able to reproduce accurately --within a large radial box-- continuum-type functions, on an energy range suitable for a number of physical applications. Using such a tool, we could provide an analytical formulation to obtain molecular photoionization observables within a one-active-electron model and a single-center approach. The reliability and efficiency of the method was shown by studying the photoionization of several neon-like molecules (NH3, H2O and CH4) [2].

Here we present also an application to the ionization of such molecules but by electron impact [3].  All the necessary transition integrals are again expressed in closed form. Differential cross sections calculated for methane compare well with other theoretical calculations and with available experimental data in asymmetric coplanar geometry.

Further exploiting the mathematical advantages of cGTO, our strategy for the continuum state can be equally applied together with a multicentric initial state description [4] and opens up for several applications, for example for ionization processes involving large molecules.

[1] Ammar A et al, 2020 J. Comput. Chem., 41, 2365.

[2] Ammar A et al, 2021 J. Comput. Chem., 42, 2294.

[3] Ammar A et al, submitted.

[4] Ammar A et al, 2021  Adv. Quantum Chem., 83, 287.   

Improving electron transport in Monte Carlo simulations using high-fidelity collision models

Current particle-in-cell Monte Carlo (PIC-MC) plasma simulation codes utilize an eclectic set of atomic and molecular cross section data to perform calculations. The availability of differential (angle and energy) resolved cross section data suited for simulation purposes is often limited, and simplified scattering models are used in situ. Direct simulation plasma codes offer a platform on which to test these more advanced models easily for a variety of purposes.

In this talk, we demonstrate the effect of differential elastic and ionization cross section models on the electron transport in swarm conditions and low-temperature plasmas. Several gasses are simulated using high-fidelity anisotropic elastic scattering models and electron ionization energy sharing models. We use a newly developed 0D direct simulation Monte Carlo (DSMC) code "ThunderBoltz", and a 1D PIC-MC code "eduPIC" to produce quality electron velocity distribution functions, kinetic reaction rates, and electron mobility and diffusion coefficients. We compare these transport parameters to demonstrate the importance of using higher fidelity differential collision models and outline scenarios where these more detailed models are necessary.