1. milkshake on July 23, 2009 12:05 AM writes...

Your posting is getting irregular? Try Activia - you will like how the story ends.

2. Anonymous on July 23, 2009 10:06 AM writes...

I prefer good ol' Raisin Bran.

3. InfMP on July 23, 2009 7:11 PM writes...

Derek you should do a post on the famous JACS NaH oxidation paper that has been proven false at totallysynthetic.com

4. bad wolf on July 27, 2009 10:22 AM writes...

Derek must have gone to San Diego Comic-Con. Nerd!

5. Jan Teller Jr on July 27, 2009 3:30 PM writes...

I love that blog, but the control experiment designed by totallysynthetic wasnt precisely top-notch...

Actually, the authors had an entry with an EW-substituted benzylic alcohol (just as TotSyn) giving a 12% yield after 30 h in exactly the same conditions detailed on the paper...so no Gate, no scandal, just bored chemists I guess...

6. Hap on July 29, 2009 9:48 AM writes...

No...TS's example was the exact duplicate of a substrate which gave 85% yield for the authors - the substrate you're comparing it to is a p-nitro.

So, no, his example was precisely correct, and the conclusion from the reaction reasonable. Thanks for playing.

7. Jan Teller Jr on July 29, 2009 10:14 AM writes...

No playing, T

The rigourosity of the scientific method should work in both ways, no?

Just trying to reproduce, in partially similar conditions, the most disfavoured hit on the list doesn´t seem to me formal enough for a specialised blog of TS´s category and repercussion.

Thanx for commenting

8. Hap on July 29, 2009 11:58 AM writes...

No - just arguing that the test case he tried was accurate. Trying more reactions would have helped to firm up his conclusion, but his conclusion wasn't unreasonable based on his experiment.

Trying a crappy-yielding oxidation would have been problematic - the rate would have likely been low, so it would have taken longer, and if the yield were lower, it wouldn't have necessarily said anything (would getting 0% under Ar when they got 12% mean that their method was incomplete, or just that the substrate sucked?) Maybe he could have tried multiple oxidations, but he would have tried ones with high yield in any case (to easily distinguish their behavior from that reported), and the only other substrate that looks readily available would be the alpha-methyl-p-methoxybenzyl alcohol. The only caveat might be that a low yield doesn't distinguish between bad technique and failure in the absence of O2 (though the reactions of others support the latter scenario, and TS has succeeded at other, more complicated work). I don't see where his conclusions or methodology were wrong, exactly.

Sorry for the snark, though.

9. Jan Teller Jr on July 29, 2009 3:54 PM writes...

No worries abou the snark though...a wee one before taking annual holidays is OK!

My point was simply that he should have, as you said, firm up better his post before doing it. We should not understimate the power of the blogosphere nowadays specially if you have one supported by the RSC.

In my humble opinion, he should have done a control experiment with oxygen an no NaH, and a try with the analogous methoxy-substituted substrate before throwing the post and perhaps shedding doubts on the papers, and extensively on the author´s reputation

10. Hap on July 29, 2009 5:19 PM writes...

The drybox experiment reported on Carbon-Based Curiosities would be the key one - if one can show that the reaction doesn't go without O2 and does in its presence, that would go pretty far in explaining the original results and making them useful. Probably can't use it on scale, but it might be OK in lab.

The lead author's explanation on RSC makes sense, but his group probably should have done a better job making sure that O2 wasn't the oxidant. The reviewers probably should have brought it up, too.