Our interest is in inner-shell processes that involve X-ray emission. Our results help to test predictions based on the independent-particle model (IPM) and to probe situations where electron correlation effects make the IPM unsatisfactory. The quantities of interest are of practical importance because they are part of the database necessary for PIXE and other X-ray emission techniques. We measure the widths of X-ray lines and of inner-shell states, and also the branching ratios between radiative and non-radiative de-excitation of the latter, viz. subshell fluorescence yields and Coster-Kronig probabilities. We have worked on relative X-ray intensities in the K, L1, L2 and L3 series, and on angular correlations and distributions of X-rays following ionizing processes. Much of this work was in collaboration with Dr. Tibor Papp. Satellites are also of interest, arising from processes such as double ionization, shake, the radiative Auger effect, and Coster-Kronig de-excitation.

Atomic level widths (K - N7)

Work was concluded in 2001 on a large project "commissioned" by the journal Atomic Data and Nuclear Data Tables. We re-assessed all direct and indirect level- and line-width measurements since 1930 (from 12 different spectroscopies), updated the data analysis where necessary, and produced a recommended self-consistent set of widths for the K-N7 levels. This is intended to replace the standard reference works done in the 1970s, e.g., Keski-Rakhonen and Krause [ADNDT 14, 139, 1974]. Our work highlights the general failure of IPM theory when Coster-Kronig transitions play a role, targets areas where experiment is needed, and demonstrates the need for many-body-theory efforts.

L subshell fluorescence and Coster-Kronig yields

Another project for ADNDT comprised a critical assessment of these important quantities, and the production of a recommended set of values. The CK process re-arranges the subshell vacancies after the initial ionization event and prior to X-ray emission, and thus the detected spectrum and the "expected" spectrum differ. To work back from the detected spectrum to the initial ionization distribution, these quantities are vital.

The three CK probabilities and the L1 fluorescence yield are poorly known and difficult to measure. In the past we did a series of coincidence measurements of the L2-L3 Coster-Kronig probability f23, showing that it fell typically 10% below the IPM prediction. We had also studied the six L subshell yields in Gd. Recently we did a new experiment to determine these quantities for the case of lead, where previous experimental results differed significantly from theory; in such experiments the fitting of the spectrum with appropriate lineshapes and correct description of the background, and inclusion of satellites, is a crucial ingredient.

K X-ray intensity ratios where the 4d subshell is open

The independent-particle model predictions of K X-ray intensity ratios accord well with measured data except in the interesting area (21 < Z < 30) where the 3d shell is open, resulting in coupling between the 3d angular momentum and the final 3p vacancy; also the de-localization of the 3d electrons in the solid causes strong enhancement of the l-forbidden K-M4,5 transition. Such effects might also be expected in the 40 < Z < 50 region where the 4d shell is open, but data are few and only one paper had previously reported an incontrovertible effect. We used our GPPV software to analyse proton-induced X-ray spectra, and observed a systematic departure of the K2/K1 ratio from the IPM value, maximum when the 4d is half-full.

L subshell ionization cross-sections for protons

Almost all measurements in this area have been interpreted using the L subshell fluorescence and Coster-Kronig yields recommended in a 1979 review. And almost all measured spectra have been fitted without recognition of the intrinsic Lorentzian shape of x-ray lines or of satellite effects. Differences between measured and ECPSSR cross-sections have led to development of corrections such as the United Atom and the Intra-Shell vacancy transfer effects. We have just concluded a series of careful measurements designed to ascertain if the differences are decreased by using our own 2002 set of recommended L subshell yields: a paper has been submitted and the results reported in it suggest a significant improvement through the use of these yields. The combination of ECPSSR-DHS theory with the 2002 Campbell fluorescence and Coster-Kronig yields provides excellent agreement with measured Lβ/Lα and Lγ/Lα ratios using 2.0 and 2.5 MeV protons.


  1. J. L. Campbell and T. Papp. Widths of the atomic K - N7 levels. Atomic Data and Nuclear Data Tables 77, 1-56 (2001).
  2. J.L. Campbell. Non-monotonic behaviour of the K2'/K1' x-ray intensity ratio in the atomic number region 40 < Z < 50. J.Phys. B: At. Mol. Opt. Phys. 34, 3543-3553 (2001).
  3. M. Kavcic, M. Budnar, (both Inst. J. Stefan, Slovenia) J.L. Campbell. A database for KL ionization satellites in PIXE. Nucl. Instr. Meth. B196, 16-24 (2002).
  4. J.L. Campbell. Measurement of the L subshell fluorescence and Coster-Kronig yields of bismuth. J. Phys. B: At. Mol. Opt. Phys. 36, 3219-3231 (2003).
  5. J.L. Campbell. Fluorescence yields and Coster-Kronig probabilities for the atomic L subshells. Atomic Data and Nuclear Data Tables 85, 291-315 (2003).
  6. B.N. Jones and J.L. Campbell. L X-ray production cross-section ratios for protons incident on high-Z atoms: a test of ECPSSR theory and newly recommended vacancy de-excitation parameters. Nucl. Instr. and Meth. B258, 299-307 (2007).
  7. J.L. Campbell. Fluorescence yields and Coster-Kronig probabilities for the atomic L subshells. Part II: The L1 subshell revisited. Atomic Data and Nuclear Data Tables (2008), doi:10.1016/j.adt.2008.08.002.