C. W. McCurdy and W. A. Isaacs,
Lawrence Berkeley National Laboratory
T. N. Rescigno, Lawrence Livermore National Laboratory
M. Baertschy and D. A. Byrum, University of California, Davis

Research Objectives

This project seeks to develop theoretical and computational methods for treating electron collision processes that are currently beyond the grasp of first-principles methods. We are developing methods for studying electron-atom and electron-molecule collisions at energies above that required to ionize the target, for calculating detailed electron-impact ionization probabilities for simple atoms and molecules, and for treating low-energy electron collisions with polyatomic molecules, complex molecular clusters, and molecules bound to surfaces and interfaces.

Computational Approach

Our approach builds on the algebraic variational formalisms we have been developing to study electron-atom and electron-molecule scattering. These approaches have now been extended to include complex optical potential interactions, a scattered-wave/flux operator formalism, and a variety of techniques based on analyticity. These techniques allow us to treat a broad range of problems, from low-energy electron-molecule collisions using elaborate variational wave functions to direct solutions of the Schr�dinger equation for simple atomic targets that provide detailed ionization cross sections. Our work uses the mathematics of complex analysis in numerical computations, and complex coordinates were the key to solving the electron-impact ionization problem completely for the first time.

One component in an expansion of the wave function describing electron-hydrogen scattering plotted vs. the distances of the two electrons from the nucleus. The large-amplitude oscillations along the edges correspond to one outgoing electron, with the other remaining in a hydrogen bound state. The oscillations in the middle represent two outgoing electrons after ionization of the hydrogen atom.

Accomplishments

We have published major theoretical studies of low-energy electron scattering by CCl₄ and BCl₃-gases that are both used in the plasma etching of silicon and for which little experimental data exits. We have also completed the first ab initio study of low-energy electron scattering that has been able to produce results that are in quantitative agreement with measured cross sections. We successfully developed a code that allows for parallel computation of the most computationally intensive part of electron-molecule scattering calculations-the collision integrals needed to describe electron-exchange effects at low energies.

We demonstrated the first complete, first-principles numerical treatment of the electron-impact ionization problem. Our computational approach allows us to extract ionization cross sections (total and differential) directly from a scattering wave function that is constructed without explicit imposition of asymptotic ionization boundary conditions. We have obtained the first results of a complete treatment of the full three-body Coulomb problem with no approximations. Our approach requires the iterative solution of very large (several million by several million) systems of complex linear equations.

Comparison between calculated (lines) and measured (dots) cross sections for electron-impact ionization of hydrogen. Each panel corresponds to a particular angular separation between the two outgoing electrons.

Significance

Electron collision processes play a key role in such diverse areas as fusion plasmas, plasma etching and deposition, and waste remediation. Electron-molecule collisions play a central role in the plasma processing of silicon chips for the manufacture of very large-scale integrated circuits. All the interesting chemistry in these plasmas is electron initiated, and electron impact dissociation of etchant gases produces the reactive fragments that perform the relevant surface chemistry. The understanding and modeling of these low-temperature plasmas is severely hampered by the lack of a database of electron-molecule collision cross sections. This project will significantly add to that base of knowledge.

Our development of the first complete method for treating electron-impact ionization of atoms from first principles solved a fundamental problem in atomic physics that has resisted solution for more than 40 years. Electron impact dissociation, attachment, and ionization phenomena also occur in the condensed phase, notably in mixed radioactive waste. The primary radioactive decay events produce showers of secondary electrons which initiate the chemistry that produces a mixture very different from the one that was present originally. The tools we are developing will make it possible to understand and model this kind of electron-initiated chemistry.

Publications

T. N. Rescigno, M. Baertschy, W. A. Isaacs, and C. W. McCurdy, "Complete break-up of a colliding quantum system of three charged particles," Science 286, 2474 (1999).

T. N. Rescigno, D. A. Byrum, W. A. Isaacs, and C. W. McCurdy, "Theoretical studies of low-energy electron- CO2 scattering I: Total, elastic, and differential cross sections," Phys. Rev. A 60, 2186 (1999).

W. A. Isaacs, C. W. McCurdy, and T. N. Rescigno, "Low energy electron scattering from BCl₃," Phys. Rev. A 58, 2881 (1998).

http://www.nersc.gov/news/newsroom/sciencecover12-24-99.php