Precision and detection limit of quality test for amorphous drug in powder X-ray diffractometry.

This report puts forward a method of powder X-ray diffractometry to estimate the precision and detection limit of the crystalline component in an amorphous drug. Cefditoren pivoxil (CP) was employed as a model drug. The major error source of the measurement at low crystal contents is shown to be the random noise in a diffraction pattern (halo pattern) of the amorphous material. For the analysis of the noise, the obstructive halo pattern should be eliminated from the observed pattern. The subtraction of the observed halo pattern from another one derived from the same material, extracts the random noise alone, although the noise is amplified by square root 2 times. The noise in the powder X-ray diffractometry was identified as the white noise. On the basis of the stochastic properties of the extracted noise and signal parameters (peak area) of CP, the relative standard deviations (R.S.D.) of the area measurements of the crystalline diffraction peaks were estimated over a wide range of crystal contents without repeated experiments. The detection limit was determined such that the crystal content at detection limit produced 30% R.S.D. of the measurements. The R.S.D. and detection limit obtained from FUMI theory were in good agreement with the results from the repeated measurements.