Comparison of theoretical conversion coefficients

Abstract

The tables of internal conversion coefficients by Rose and by Sliv and Band were compared and the deviations plotted for all 17 371 cases where the parameters in the two sets of tables match. The mean value of (α R −α SB ) α SB is nearly zero, but the standard deviation is 17%. The deviations (excluding those caused by differences in models) in the K-shell are small (less than 2%) except for two regions where logical explanations exist. The deviations are large and systematic throughout the L-subshells. Many of the deviations are due to defective interpolations as a function of Z. Empirically screened (pure Coulomb) calculations, which include accurate finite size corrections, indicate inaccuracies in the tables by Rose which appear to be associated with finite size effects. These calculations also provide systematic information concerning a cancellation which greatly reduces the magnitude of conversion coefficients in certain regions of the L I subshell for electric transitions. The experimental anomalies recently reported for L-subshell ratios for pure E2 transitions are in such a region and are discussed. There are often large discrepancies between the two sets of tables at low energy. Two regions were found in which large inaccuracies apparently exist in the tables by Sliv and Band. The empirically screened program was found to be a means of accurately interpolating between values calculated with more sophisticated programs.