B.R.S.M. When all you have is a hammer everything looks like a nail


What’s Wrong With This Picture 1: Gibberellic Acid?

For people who say that inflation of yields is a new thing... think again. Corey's synthesis of gibberellic acid is otherwise really quite good.

Comments (10) Trackbacks (0)
  1. ……Did someone not wait for their product to dry perhaps?

  2. Sounds like somebody should grab Hudlicky’s paper (http://bit.ly/zV2c8q), jump in a time machine, and glue it to the chemists’ hoods.

    My grad advisor always said that no matter how good our reactions ran or our ee’s matched the lit, we would never say “100%” or “>99% ee.” Precious few reactions, and certainly even fewer in academia, can reliably reproduce those numbers.

  3. in the defense of the paper, alcohol addition to THP followed by in situ reduction with borohydride might be reasonably expected to give a nearly quantitative yield. (A perfect regio and stereochemistry again is not so shocking because one keto group is actually a vinylogous ester and cis decaline based ketonoe famously reduce with very high endo preference). The quantitative MOM protection is somewhat fishy though.

    there is an old joke called Corey 1,2 rearrangement. Say your reaction will give maybe 19% of isolated product on a good day. And the Chief observed for the fifth time during the last five meetings that this reaction should work better. So you perform the famous 1,2 rearrangement, report 91% yield in your JACS paper, burry the b.r.s.m./stereoisomer separation-re-equilibration details in supplementary, finish your postdoc and move to a pharma senior scientist position…

  4. I had heard that some people took NMRs in carbon tet (which would have made sense for that era) and if all of the solvent was not removed it would “enhance” the yield while being NMR-silent.

    One of my professors in undergrad worked on that synthesis – Danheiser’s article (in Tactics in Total Synthesis? – book) is pretty entertaining.

    • My first PI told my the same thing (only d-chloroform instead of C tet) specifically about Corey’s group. I must be mentioned though that he did a postdoc in Barry Sharpless’s group.

  5. Borax (Na2B4O5(OH)4.8H2O) is made from kernite (Na2B4O6(OH)2.3H2O). The yield is about 140% (by weight) – kernite is basically dehydrated borax. Can anyone do any better?

  6. OMe should be replaced by OEt after the second step?

  7. I don’t deny that inflation of yields occurs all too often, but it turns out that what we have here is not really a case of inflated yields, but rather an error by Corey in writing up the work.

    If you check the relevant Harvard Ph.D. thesis, you will find in the text that the yield for these three steps is given not as 100%, but rather as 98, 97, and 92%. The Experimental Section indicates that the crude product of each reaction was “used in the next step without purification”. The weight of the crude product in each of these three reactions is reported as 100 or greater than 100% relative to theoretical. In each case the Experimental reports that a sample of the crude product was purified by column chromatography (to provide the yields cited above for each step and to generate samples for characterization).

    In the Corey synthesis, after the MOM protection shown in the equation above, the ketone is reduced and the product is exposed to MsCl and then quenched with aqueous bicarbonate solution to yield an enone (compound 10 in Corey’s paper). The thesis experimental reports an overall yield of 77% for the conversion of the quinone shown above (compound 6 in Corey’s paper) to enone 10 (i.e., without purification of any intermediate products). Though this represents a rather high yield for each step in the sequence, it is not at all unreasonable.

    Corey’s paper does in fact correctly report the formation of “10 in 77% overall yield from 6”. Obviously, no yields should have been mentioned for the first three steps in this sequence, or the yields should have been qualified as being for crude products (though of course numbers for “crude yields” are pretty meaningless).

    To summarize, the actual experimental work in this synthesis is completely sound, and the overall yield is correctly reported in the paper, without “inflation”, for the sequence in which the above three steps are a part. There are plenty of other cases in the literature where “yield inflation” is probably involved, but this is not really one of them.

Leave a comment

six − 4 =

No trackbacks yet.