Sunday, January 6, 2013

Turning Wishes into Horses

A fantastic summary crossed my desk today, courtesy of the RSC's Chemistry World. It profiled a new reaction from the Milstein group, just reported in Nat. Chem., that just looked so tempting: oxidation of a primary alcohol, using water as the "O" source, with hydrogen gas as the lone byproduct.*

And thus began what my brain usually does, "Science Wish Syndrome." Not familiar? Here's the stages:

1. Unbridled Enthusiasm - "Holy cow, that reaction looks fantastic! I'm sure it will produce 20 new analogues / help me wrap up my thesis early / solve world hunger. I can't wait to try it!"

2. Skepticism / Doubt - "Well, Blogger X tried it, and couldn't reproduce the yields. And the Supporting Info says it only works with 100 mesh hyper-refined silver powder sourced from Argentina. Still, it's in Nature / Science / JACS / Angew Chem, so it must work, right?"

3. Despair - "I ordered all the reagents, and used my last 0.3 mg of shootmenowicene in the test reaction. I barely got 10%, and it's a mixture of stereoisomers! Who the heck let this into this journal?"

Readers, does this ever happen to you? Modern chemistry publication, with ASAP format, DOI codes, PDB uploads, and overall rush to publish before the guy on the other side of the pond, seems forced somehow, and doesn't always live up to the glowing reviews crowed about in press releases.

Over at Not the Lab, Vinylogous has suggested that we compile a sort of post-hoc online review journal, playfully named "Blog Syn" (after 800-lb-gorilla-in-the-room-for-synthetic-methods Org. Syn.).

Well, how might something like this work? I'd assume that the community would have some sort of mechanism to suggest papers, say, for when too-good-to-be-true chemistry like the NaH 'oxidation' comes 'round. A central aggregation point, perhaps cultivated by an "Editor / Blogger" would then issue requests for duplication. To ensure results were processed in a timely fashion, a deadline would be suggested (2 weeks?) and the "Duplicators" would have to authenticate that they tried to cleave to the original conditions as best they could. A digital "Stamp of Approval" might then follow the reaction through the online world, indicating, much like Org. Syn., that the reaction works in others' hands.

Readers, would anyone like to be part of such an effort? Methinks the #RealTimeChem crowd could get excited about this. Yes, I know everyone's busy, and this isn't the first thing you think of when there are proposals to write, classes to teach, or conferences to attend. But the literature will only be as good (or as bad) as the folks who contribute, and we desperately need more quality control.

*Update - It occurred to me (too late!) that I sound like I'm condemning the Milstein reaction, which I have not tried, and have no reason to suspect won't work as of this writing. I just wanted to use it as a backdrop for seemingly "perfect" reactions that end up not living up to expectations. The reaction, as reported, needs some scope exploration, so only time will tell if it's as widely applicable as I'd hoped.


  1. This might also be great to teach 1st years or rotation students a lesson about believing preps, and to give them some lab experience before ruining otherwise precious material.

    1. A tough (but important) lesson to learn, as early as possible!

  2. I think this sounds like a good idea. At the very least, a curated, searchable collection of online discussions related to literature (i.e. pop in the DOI and see what people have said, tried, etc) would be very valuable--it would be easy for people to search before ordering reagents to try a procedure.

    A concern would be, of course, trying procedures: it would be important, I think, to get some PIs to jump on board so as to not provoke anger when they find out their grad student has been chewing up reagents and time for a non-research pursuit. Of course, some people have autonomy to try things, some don't. But there are PIs who have their students participate in Org Syn, so I think it's feasible.

  3. For a while, Alison Frontier's Not Voodoo had a section for papers that were irreproducible -- naturally, the Evans aldol ended up on there, which I found odd (and so did commenters.)

    I don't really know what happened to it -- I can't find it when I went looking.

    "Blog Syn" sounds like a decent idea; one could imagine that a Blog Syn contribution could be a required part of a candidacy exam.

    1. Evans aldol needs good air-free technique/pure dry solvents and the Lewis acid reagent Bu2BOTf craps up if stored improperly. So it could have been one desperate grad student in a bio- lab trying to reproduce a "surefire" transformation without anyone to tell him that one has to check for presence of MeOH in DCM from a Fisher can (CaH2 or Al2O3 solvent purification column wont remove 0.5% of MeOH in DCM). And so on

      here is the page with the hard to reproduce experiments