'Pure by NMR'?

A paper at the end of 2008 by Steve Davies et al (Org Lett, 2008, 10 (23), 5433) raised the question of purity assessment by proton NMR.  In the early phases of chemical investigation or development quantitative methods of analysis may not yet be available (eg HPLC) so NMR is the routine fall-back and we often see reported in the literature, or indeed internal reports, that a material is ‘pure by NMR’.  Davies et al describe a new NMR methodology, illustrating the technique with the pair of oxazolidinones shown below wherein they could quantitate diastereoisomeric ratios up to 1000:1 (ie 99.8%de).

A standard technique to estimate the w/w assay of a sample is to collect the NMR with an internal standard added and integrate the signals for standard vs sample.  Davies et al have used this principle but used the 13C-H satellite signals of the major isomer effectively as the internal standard, taking the integral of this satellite in the major isomer and comparing it to the corresponding 12C-H signal in the minor isomer.  Since the abundance of 13C is 1.108%, then there is always a fixed amount of the ‘internal standard’ present.  There are some specific requirements in order to be able to apply the technique, but as a potentially simple method to screen samples or for use prior to development of a suitable HPLC method, this could provide a great start-point of laboratory chemists.  The authors comment that it is apparent that when the height of the 13C-H satellite resonances associated with the major diastereoisomer are greater than that of the 12C-H resonances of the minor diastereoisomer, the diastereoisomeric ratio is >180:1 (>98.9%), thus allowing a speedy approximate assessment to be made.