Low energy, large angle electron impact spectroscopy.

Abstract

Previous theoretical and experimental work has indicated that at low incident energies, singlet-triplet transitions may take place via a spin exchange mechanism and that cross-sections for these processes decrease more slowly with angle than do those for allowed transitions. Low incident energies and large scattering angles are therefore indicated when studying optically forbidden transitions. Consideration of the signal available at large angles suggests the use of an analyser capable of accepting 2 π in azimuth and this approach was pioneered by Kuppermann and Raff(l963) who used a retarding field energy analyser(R.F.E.A.) with cylindrical grids. In the present work, a study of the factors affecting resolution and peak shape in a R.F.E.A. leads to the choice of spherical grids while a split scattered electron collector is used to determine the angular dependence of relative cross-sections. Improvements in the electron optics and the use of a clean U.H.V. environment free from magnetic fields leads to a spectrometer capable of operating at incident energies of 10 to 50 eV over the angular range 15° to 165° with a resolution of about 400 to 500 meV. Spectra may be obtained simultaneously on three collectors using a data logging system with automatic scanning. Spectra obtained for N2, CO and C2H4 show that at incident energies close to threshold, and large scattering angles, singlet-triplet transitions can, in favourable cases, be more strongly excited than singlet-singlet transitions. It is found, in general, that the intensity ratio for a triplet state relative to a reference singlet state increases rapidly with increasing scattering angle at a fixed impact energy and this may serve as a test for a triplet state. This ratio is not, however, so dependent on incident energy in the 20 to 50 eV range and, in general, only increases noticably when the incident energy is within 10 eV of threshold. As regards the controversy arising from the discrepancy between the results of Kuppermann and Raff(1963) and Simpson and Mielczarek(1963) for ethylene, the present work gives results consistent with electron trap spectra and large angle, energy loss spectra using dispersion analysers, showing only peaks at 4.4 eV, 8 eV and 9.5 eV with no indication of the so-called "mystery band" at 6.5 eV.