Developing Cross Section Sets for Fluorocarbon Etchants

Abstract

Successful modeling of plasmas used in materials processing depends on knowledge of a variety of collision cross sections and reaction rates, both within the plasma and at the surface. Electron-molecule collision cross sections are especially important, affecting both electron transport and the generation of reactive fragments by dissociation and ionization. Because the supply of cross section data is small and measurements are difficult, computational approaches may make a valuable contribution, provided they can cope with the significant challenges posed. In particular, a computational method must deal with the full complexity of low-energy electron-molecule interactions, must treat polyatomic molecules, and must be capable of computing cross sections for electronic excitation. These requirements imply that the method will be numerically intensive and thus must exploit high-performance computers to be practical. We have developed an ab initio computational method, the Schwinger multichannel (SMC) method, that possesses the characteristics just described, and we have applied it to compute cross sections for a variety of molecules, with particular emphasis on fluorocarbon and hydrofluorocarbon etchants used in the semiconductor industry. A key aspect of this work has been an awareness that cross section sets, validated when possible against swarm data, are more useful than individual cross sections. To develop such sets, cross section calculations must be integrated within a focused collaborative effort. Here we describe electron cross section calculations carried out within the context of such a focused effort, with emphasis on fluorinated hydrocarbons including CHF3 (trifluoromethane), c-C_(4)F_(8) (octafluorocyclobutane), and C_(2)F_(4) (tetrafluoroethene).

Files

Additional details