Autoionization electron impact excitation

Autoionization spectra resulting from specific resonances can be obtained by electron-electron coincidence measurements (Haak et al. 1984 Ungier and Thomas 1983, 1984, 1985). To associate a fr.rgmentation pattern with a particular core hole excited state and a particular autoionization or Auger decay channel, a double-coincidence experiment must be done using electron impact excitation. The energy of the scattered electron must be determined, the energy of the emitted electron must be detennined, and the ions produced in coincidence with these two events must be determined. The difficulties inherent in these kinds of experiments have been aptly summarized by Hitchcock (1989), If you can do it by photons, don t waste your time with electron-coincidence techniques. ... 

Another possibility is the analysis of ejected electron spectra due to electron impact excitation of autoionizing states via a negative-ion resonance. One example, investigated by van de Water and Heideman, is the excitation of He(2s ) 5 via the negative resonance He (2s2p ) D. Several processes can lead to the same final state ... 

Additional details on some of these methods are described in other sections of this review. Attempts have also been made to determine excited-state populations in single-source mass-spectrometric experiments from an analysis of ionization efficiency curves.38ad There are several difficulties in applying such methods. For instance, it is now known from photoionization studies that ionization processes may be dominated by autoionization. Therefore, the onset of a new excited state is not necessarily characterized by an increased slope in the electron-impact ionization-efficiency curve, which is proportional to the probability of producing that state, as had been assumed earlier. Another problem arises because of the different radiative lifetimes that are characteristic of various excited ionic states (see Section I.A.4). 

The interaction of an electron with a molecule is described as a collision or impact, although the electron is so small that there is no collision in the usual sense of the word. The collision process may be termed elastic (the electron is merely deflected), inelastic (energy is transferred from the electron to the molecule), and superelastic (energy is transferred from the molecule to the electron). Electron-impact ionization is an example of an inelastic collision. The energy imparted to a molecule during an inelastic collision can lead to rotational, electronic, and vibrational excitation with or separate from ionization. Further, multiple-electron excitation can occur followed by autoionization, and the latter process has been shown to lead to a substantial fraction of total ionized species in many cases (S. Meyerson et al., 1963). Thus, an electron of energy 20 eV may lead to any of the above excitations of a molecule. The gas pressures used in a mass spectrometer and the density of electrons in the electron-beam are such that multiple electron-molecule interactions leading to ionization are improbable. 

About SEFS as a spectroscopic feature, it can be said that this structure is a manifestation of effects of the strong electron-hole correlation (with a hole on the atom core level) upon excitation of the atom by an electron impact. As opposed to the classical autoionization effects (Fano effects), where the presence of a resonance level is assumed, in the present case a strongly correlated state is observed... 

Figure 5. Autoionization spectrum from He, excited by electron impact. The line shapes are strongly influenced by interferences between direct ionization and autoionization. (From Ref. 10.)...

Figure 17. Angular distribution of autoionization electrons from Ne (2p 3s ) D, excited by He impact. A-A shows the distribution in the scattering plane, B-B in the plane perpendicular to the He -beam direction, and C-C in a plane perpendicular to the scattering plane, tilted by 45 with respect to the He -beam direction. Solid curves represent the result of a fit calculation. Parameters obtained from the fit are used to calculate the complete distribution, which is shown in the three-dimensional view. 

In their energy spectra of H ions emitted after impact of p and e on Hj, Edwards et al. [6.10] were able to identify a small contribution stemming from double excitation to the states iTg,, or, followed by autoionization. The total cross section for these processes was found to be a factor of 3 larger for electron impact than for proton impact at 0.35 MeV/amu, decreasing to a value of 2 at 3.5 MeV/amu. This result is in qualitative disagreement with the findings discussed above for the He target. Whether this discrepancy is due to peculiar circumstances connected with the molecular structure of Hj is unknown at present. 

In their measurements of electrons emitted after double excitation of the (2p ) D and (2s2p) P states of helium by 1.84 MeV/amu electron and proton impact, Pedersen and Hvelplund [6.23] observed that the yield of autoionized electrons was considerably smaller in the forward direction for electron impact than for proton impact. The authors do not believe that this can be due to a postcollision effect like that seen by Skogvall and Schiwietz [6.29] since the projectile electron velocity in their case is much larger. They suggest that the reason is instead a difference in the excitation of the doubly excited states, combined with the effect of interference between the amplitudes for ionization after double excitation and direct ionization. 

This experiment may be regarded as the forerunner of mass spectro-metric appearance-potential determination in that both are threshold techniques, that is they depend on slow variation in the energy supplied by the impacting electron until a change in the electron-molecule interaction is observed. Thus, just as the Hertz experiment did not distinguish between excitation and ionization potentials, mass spectrometric appearance potential measurements are subject to similar ambiguities in interpretation as between ionization and autoionization. 

Since the photoabsorption measurements of Madden and Codling autoionization processes have been investigated by various methods. Excita-tion has been initiated by electrons, " by heavy particles, or by beam-foil interaction. Whereas the number of states that can be excited by photon impact is limited by selection rules, this limitation is less stringent for electron collisions, especially at low impact energies. For ion-atom or atom-atom collisions it is possible to provoke or suppress the excitation of certain types of autoionization states by careful selection of the collision partners. ... 

The situation is more complicated if several autoionizing states in He are excited by low-energy ion impact such that postcollision interaction leads to interferences between transitions from these states. As an example Figure 20 shows coincident electron energy spectra from Li" + He collisions. 

---