The following is quoted in Kanare, p69. It is a salutary lesson in recording detail. Never mind the details of the chemistry - note that the small point about the solvent transfer, omitted in the report originally, makes all the difference to the experiment and its reproducibility.

Sir Geoffrey Wilkinson was at Imperial College, University of London, and co-author of Cotton and Wilkinson, a standard University inorganic text.

How much detail should be recorded in your notes? Could another scientist who is competent in your field pick up your notebook and repeat your work solely from the written description without additional explanation? If the answer is yes, then you are doing a good job. Too many details are better than for you to assume that a future reader (perhaps you!) will know all about your work. A story told by Nobel Laureate, Sir Geoffrey Wilkinson, illustrates this point:

Mr. A. J. Shortland made the first synthesis of hexamethyltungsten, one of the biggest breakthroughs in transition metal organometallic chemistry, by a method involving interaction of methyllithium with WC1₆ in ether. This was described in J. Chem. Soc. Dalton Trans. 1973, 872, as well as in his Ph.D. thesis in which we had stressed the necessity for rigorous oxygen-free conditions.

It came to my notice subsequently that some others had been unable to reproduce the work. Consequently I had a new student, Mr. L. Galyer, try to repeat the work as described and he also failed. Since Dr. Shortland now had a teaching job at Dulwich College in London, I asked him to come back and demonstrate - which he willingly did. The crux was that he had degassed the petroleum used to extract the evaporated reaction mixture separately in another flask by bubbling nitrogen through it. Instead of transferring the petroleum to the reaction flask via a steel tube and serum cap technique, he actually removed a stopper and poured the petroleum into the flask. This immediately partially resaturated the solvent with oxygen - irreproducibly of course, depending on the pouring path length, flow of nitrogen out of the joint from the reaction, etc.

We are pretty sure that the initial reaction excess LiMe + WCl₆ gave a lithium alkylate anion of the type Li₂[W(IV)Me₆l with reduced tungsten and that oxygen was required to reoxidise this to WMe₆.

We had a similar experience with synthesis of ReOMe₄ from ReCl₅ (J. Chem. Soc. Dalton Trans. 1975, 607) in which we utilized this observation.

If it had not been for Shortland's sloppy technique we would doubtless have abandoned work on high oxidation state methyls.

Back