EXTINCTION CORRECTIONS FROM EQUIVALENT REFLECTIONS

Corrections for secondary extinction evaluated from the diffraction in tensities for equivalent reflections with different path lengths provi de an independent check on values that minimize differences between ob served and calculated structure tactors. Comparison of equivalent inte nsities also avoids any extinction-parameter bias, which originates in correlation of the extinction corrections with bonding-electron contr ibutions to X-ray structure factors. Corrections from the comparison o f equivalent reflections for several X-ray diffraction studies on smal l crystals of ionic compounds are markedly less than those that minimi ze differences between observed and calculated structure factors. The discrepancies that originate in extinction-parameter bias are exacerba ted by the unfavourable form of the statistical distribution function for the residuals when differences between observed and calculated str ucture factors are minimized Analysis of intensities for equivalent re flections, although more demanding experimentally, provides least-squa res residuals closer to the normal distribution required for reliabili ty in nonlinear least-squares processes.