Letter to the Editor

Need for elemental analysis

Dear Editor

Lachlan Cranswick’s excellent review of the IUCr Teaching Commission Newsletter prompted me to look up the original. This took a bit of finding as the published URL is incomplete, but that’s another story. It is Anthony Linden’s article on “Crystallographic Data Validation - Ten Years On” that has driven me to put fingers to keyboard.

As an analyst as well as a crystallographer it was immediately obvious to me that had ANY of the compounds discussed by Anthony Linden been subjected to elemental analysis, the misidentification of imine as carbonyl would have been immediately apparent. The nitrogen analysis would have been 6 or 7 % out - in some cases, a third of the total nitrogen expected wasn’t nitrogen. CHN analysis can be done on 2 mg of sample - less than 2 cubic millimetres - not very much at all. Proper chemical characterisation of compounds presented for structure determination is essential.

I would hope that papers describing crystal structure determinations will not be accepted for samples which are not adequately characterized chemically. What is “adequately characterized”? That must be decided for each case, jointly by author, editor and referees.

David Beveridge