Crystallization of trehalose dihydrate (C12H22O11·2H2O) was induced by storing tablets of amorphous anhydrous trehalose (C12H22O11) at 65% RH (RT). Our goal was to evaluate the advantages and limitations of two approaches of profiling spatial distribution of drug crystallization in tablets. The extent of crystallization, as a function of depth, was determined in tablets stored for different time-periods. The first approach was glancing angle X-ray diffractometry, where the penetration depth of X-rays was modulated by the incident angle. Based on the mass attenuation coefficient of the matrix, the depth of X-ray penetration was calculated as a function of incident angle, which in turn enabled us to "calculate" the extent of crystallization to different depths. In the second approach, the tablets were split into halves and the split surfaces were analyzed directly. Starting from the tablet surface and moving toward the midplane, XRD patterns were collected in 36 "regions", in increments of 0.05 mm. The results obtained by the two approaches were, in general, in good agreement. Additionally, the results obtained were validated by determining the "average" crystallization in the entire tablet by using synchrotron radiation in the transmission mode. The glancing angle method could detect crystallization up to ∼650 μm and had a "surface bias". Being a nondestructive technique, this method will permit repeated analyses of the same tablet at different time points, for example, during a stability study. However, split tablet analyses, while a "destructive" technique, provided comprehensive and unbiased depth profiling information.
Naveen Thakral, Hiroyuki Yamada, G. Stephenson