How low levels of amorphous content in an API can you detect?

According to the literature, various methods makes it possible to detect less than 1% amorphous material in a crystalline sample using various methods.

However, it is not possible to define a “general” limit for the detection level as it all depends on the sample in question: If large differences in properties exist between the crystalline sample and the amorphous form it is, quite obviously, easier to determine the content than if the differences are very limited.

Our approach is to combine different methods such as FTIR and DVS (Dynamic vapor sorption) (XRD is not ideal for quantifying amorphous material). In many cases the difference between the FTIR spectra of the crystalline and amorphous material is quite pronounced, and in these cases, the task is to select the optimal regions, prepare reference standards – and perform least square fit. This approach usually allows detection of around 1% of an “impurity” such as the amorphous form. Further, DVS is performed on the same standard references. It is well know that amorphous material most often adsorb more water than the crystalline counterpart.  Thus, the water uptake will be directly related to the content of amorphous material.

From the literature, it further seems that fast DSC has given quite good results. By this method, the glass transition – when amorphous material changing into crystalline upon heating – is determined.