Data requirements for the reliable use of atomic pair distribution functions in amorphous pharmaceutical fingerprinting

Purpose
To determine the optimal measurement strategy for fingerprinting condensed phases of pharmaceutical systems using atomic pair distribution functions (PDFs) obtained from data collected using several types of x-ray diffraction instruments.
Methods
PDFs of crystalline and amorphous-phase molecular systems derived from data accessible to copper-, molybdenum-, and silver-anode laboratory sources were compared to one another and synchrotron data using qualitative and quantitative methods.
Results
We find that reliable fingerprinting is still possible using silver and molybdenum laboratory sources, but data from copper anode laboratory sources are unreliable for fingerprinting, yielding ambiguous and potentially incorrect results.
Conclusion
The ambiguities make data measured using low energy x-rays unsuitable for fingerprinting active pharmaceutical ingredients and small molecule systems, and, in general, copper anode diffractometers are undesirable for this purpose; however, laboratory x-ray sources with either Mo or Ag anodes are well suited for this application.