IEEE Transactions on Nuclear Science, Vol. NS-26, No. 1, February 1979 Heavy-Ion-Induced X-Ray Spectranetry for Chemical Analysis* R. L. Watson, J. A. Demarest,t A. Langenberg, F. E. Jenson, J. R. White and C. C. Bahr Cyclotron Institute and DepartTent of Chemistry, Texas A&M University, College Station, Texas 77843 High resolution x-ray spectranetry in canbination with excitation by heavy-ion baibardment provides a number of prospective ways of obtaining information relating to the chemical state of a substance. The relative intensity distribution of the Ka x-ray satellites arising from multiply ionized atams is sensitive to the chemical environment and in same cases might be profitably employed to distinguish between various chemical species. At resolutions sufficient to discern features of the multiplet structure associated with the angular mametum cou- pling of unpaired electrons, it is found that the multiplet intensities are also sensitive to the environment and are therefore capable of providing additional chemical information. For elements below phosphorous in the periodic table, the Ka satellites overlap the K absorption edge, thereby providing a convenient and accurate means of measuring K binding energies. In this way, chemical shifts equivalent to those obtained by ESCA may be determined. Introduction It has long been known that Ka x-ray satellites are much more sensitive to the chemical enviroment than are the Kal,2 diagram lines. In particular, the I<Ll satellites, originating fran atcns having a single K- plus a single L-shell vacancy, have been found to exhibit energy shifts two to three times larger than the parent doublet and relative intensity variations of a factor of two for various aluminum compounds.1,2 However, the utilization of this chemical sensitivity for analytical purposes has not received much attention, primarily because satellite lines are only weakly excited with the traditional photon fluorescence or electron bombardment methods of x-ray production. When heavy ions are used as a means of producing K-shell ionization, the Ka satellite lines daniinate the K x-ray spectrum, and hence it becomes possible to examine chemical effects on x-ray satellite struc- ture in considerable detail. Recent studies in our laboratory have revealed a new kind of effect which causes a redistribution of intensity fram one satel- lite group to another,3,4 and in addition have extended the investigation of intensity variations among the individual multiplet component5 (which aomprise ea-h satellite group) to the KL and higher satellites. In this presentation, the general features of the effects of chemical environment on Ka x-ray satellite intensities are described. Environmental Effects on Ka Satellite Group Intensities General Considerations The effect of chemical environment on the Ka satellite group intensities may be qualitatively understood by recalling the various ways in which a particular vacancy configuration may decay, as is illustrated in Fig. 1. Consider a configuration involving one K-shell vacancy and n L-shell vacancies, produced as a result of an ion-atom collision. Then, following the collision, two possibilities exist; either the KLnJ state will decay to an LP, L+l, or Ln±2 state via a K-Auger or K x-ray transition, or it will decay to a KrP-1 state via an i-Auger or L x-ray transition. If the initial state decays directly by AUGER K L X , (n- I), (m+2) KLL AUGER nIAT. KLM~~~~~~~~~K AUGER KMM AUGER (n+2),m a X-RAY (nfl),(rntl)0 / p<,B X-RAY (n+l),m On+ln(mF2)4 n,(m+ ) L Fig. 1 A schematic diagram depicting the various decay channels for a ALn vacancy configuration. the emission of a Kac x-ray, this x-ray will contribute to the intensity of the KLn satellite group. If, on the other hand, Ka x-ray emission is preceded by one or mere i-vacancy filling processes, this x-ray will contribute to the intensity of one of the lower order satellite groups. It is to be expected that the relative probabili- ties of the various K- and I-vacancy filling processes will depend on the local valence electron density near the atcrn involved, especially in the case of second and third row elemnts where the valence electrons are in the L- and M-shells, respectively. In addition, our measurements3'4 as well as those of Hopkins et al.6 have shown that electron transfer froa neighboring atcms also plays an important role in detemining the satellite group intensities. These processes are tenred inter-atomic transitions (IAT) or crossover transitions, and they are expected to contribute mainly to the filling of L- and M-shell vacancies in second and third row elements, respec- tively. Third Row Elements An example of the effect that caopetition between K- and L-vacancy filling can have upon Ka satellite group intensities is shown in Fig. 2 where spectra for the gases SiH4 and SiF4, and for solid Si (obtained using 32-MeV oxygen ions) are capared. The most notable feature in this comparison (aside fram the peak broadening in SiH4 which is due to an instrumental effect4) is the large enhancement of the intensities of the higher-order satellites in the SiH4 spectrum. It may also be noted that the SiH4 satellite groups display sizeable shifts to higher energies as measured relative to those of SiF4 and solid Si. This latter observation indicates that a large fraction of the M- shell electrons are removed in the heavy-ion collision. The enhancement of the higher-order satellite intensi- ties in SiH4, then, is a consequence of the fact that in this essentially free-atam case there is no source 0018-9499/79/0200-1352$00.75(E)1979 IEEE