Quote:
Originally Posted by ernie
Honestly, for many organometallic compounds it is almost impossible to get a good elemental analysis. Even with combustion aids they will often come out too low on C and H. Typically, one would note this by mentioning, for example, we used the same crystals from which the structure was determined and several combustion aids but couldn't get a number to match theoretical. Also, in many cases if you were to do an error analysis on the technique for a specific compound you will come up with error bars that are larger than the degree of error the publication accepts. I'm positive I've published compounds where we quite simply couldn't get the elemental to work out despite knowing it is clean material without impurity.
With the abundance of other data collection techniques (multi-nuclear NMR, IR, Raman, Mass Spec, x-ray structure etc) the elemental analysis isn't of great importance provided the evidence from the other techniques is clear. Elemental analysis in this day and age is quite simply unnecessary in the large majority of cases.
That said, falsifying such data is obviously a big no no if this was indeed done. On the other hand the mode of operation for many grad students, post-docs and professors is to submit samples for elemental until you get one whose numbers are "correct". In the real world of quality control you can't really do that and it honestly amounts to the same thing as making the numbers up.
In my graduate lab we didn't resubmit unless we did the test different (i.e. combustion aid) or re-worked the compound in some manner. Can't say all my colleagues in the lab followed this protocol.
In general data traceability in academic environment is woeful.
|
Completely agree about EA mostly being a dubious diagnostic tool, but I think it still has an important role in many cases.
That being said, when I read that paper I laughed until I cried. I have a couple little errors in some of my ACS supp infos, but nothing like that.