Thread by @dasingleton

1. Ok, this is the thread to end all threads, or possibly my career. I am about to trash a new paper in @sciencemagazine, its editors, its reviewers, C&E News, others who stated or wrote adoring commentaries, anyone who looked at the paper or its commentaries without seeing

2. anything wrong, the often sloppy thinking of organic chemists and the increasingly irresponsible way we teach new ones. I will however try my best not to trash the authors themselves. When I was an editor, I rejected without review a paper by an NAS member that had

3/ similar-level errors, so anyone can make mistakes. Their reaction is still cool. I will however have, er, questions about the confluence of scientific and non-scientific issues that might have influenced this paper.

4/Why here? Because it touches on issues that affect science and society more broadly. Because corrections of basic science have no good venue. Because there is no payout for fixing the literature, though one would never lack things to do. It’s a brave new world, for good or bad.

5/Why bother? Because @sciencemagazine is important. When The @Tennessean publishes that “Primary cause of global warming is force of gravity” or The @WSJ publishes “The Sea Is Rising, but Not Because of Climate Change” it is bad, for humanity, but mitigated because sensible

6/people recognize that the general media often gets things wrong and is at least partially self correcting (yes in the case of the @Tennessean, no for the @WSJ). One sifts the good from the bad. The same is true of academic journals. A few months ago the twitterverse lost its

7/ mind over fake papers that got published in the “Grievance studies affair.” What was lost in the outpour is that awful papers are regularly published in the lower-level journals of chemistry or physics or in any academic discipline, and it is ok. I think of it as the training

8. of researchers, the next step beyond lab reports, and a way to keep researchers with weaker training or resources in the game for later, and people get better. There are sometimes tremendous gems in these journals, and a basic part of PhD training is learning to judge papers

9. and sift the good from the bad.

But it is different with our best journals. I would rather see a hundred more papers on dog rape at parks in obscure journals than see one more paper in Science that ignores a hundred years of chemical structural understanding.

10 /In an age when the anti-science movement is killing people, in multiple ways, we need for some scientific journals to be standard bearers, as rock-solid as humanly possible. When a Lamar Smith criticizes something in Science, we need for that to be manifestly ridiculous.

11/ Mistakes will always happen. I have previously reported work correcting a Science paper. The paper’s errors were subtle, so that is fine. My only qualm is that I could never come up with a non-bothersome explanation as to why the original paper was accepted in Science.

12/So how is it that Science published something that is fully the organic chemistry equivalent of a flat earth? It must have received great reviews. Dear Reviewers: what were you thinking? Dear Editor: If this had come to me back when I was an editor for JOC, I would have

13/ absolutely positively rejected it without review. But then, I am an organic chemist so I guess that helps.
Science then also published a glowing commentary on the paper. And then C&E News wrote it up, with more glowing comments. And people here tweeted it out.

14/ Ok, I need to explain the chemistry at both a basic level and an advanced level. At a basic level, the paper reports a mechanism that would be unacceptable in any undergraduate organic course. Although the paper’s reaction happens whatever the mechanism, it should be clear

15/ that the mechanism was key to its publication. The mechanism is emphasized in the paper’s title, in the heart of the paper, and in every commentary. This is as it had to be to get published in Science and get press, and I could imagine it affecting patent rights. That is

because the same reaction could already be accomplished by a free radical mechanism. The new conditions and catalyst appear to be more efficient, but optimized efficiency alone does not get you patents and does not get you into Science. The paper needed the impossible mechanism.

16/ because the same reaction could already be accomplished by a free radical mechanism. The new conditions and catalyst appear more efficient, but optimized efficiency alone may not get you patents and does not get you into Science. The paper needed the impossible mechanism.

17/ Now, at the undergraduate organic level we teach all sorts of things that are gross oversimplifications or just plain wrong. This is true. But that does not mean that we are wrong when we teach students that methyl cations and monovalent oxygen cations are too awful to draw.

18/ Instead, these ideas reflect a deeper truth, that these cations are very high in energy. That means that they can only be formed under very specialized conditions, mostly the gas phase, and only when there is simply no alternative, because anything else will be better.

19/ Let me fill in a few of the advanced details. The solvent is sulfuric acid, and while it isn't a strong base, it does have a higher proton affinity than SO3 (on SO3's oxygen, not sulfur, of course, because the sulfur is an electrophile, not a nucleophile) by about 100 kJ/mol.

20/ That means that if we actually did form a CH3+ under the reaction conditions, if if it somehow managed to make a choice during its half picosecond lifetime (being generous here), it would react with the sulfuric acid, not the SO3. Once the methyl is on the sulfuric acid, it

21/ is never coming off because a proton will flow to a lurking bisulfate megacluster. The methyl sulfate, once formed, would just sit there and laugh at you.
If by some miracle a methyl cation did manage to react with an SO3, reaction at the oxygen give an ok cation that I

22/ will estimate is 50-100 kcal/mol better than the cation you get if the bond is made to the sulfur. Any structure with a sulfur-bound methyl would be much higher in energy than methyl cation, I will guess again 50-100 kcal/mol (anyone want to bet money against me?), because an

23/ O+ is just simply that much worse than a C+.

And while it is a relatively minor issue, the peroxide catalyst can't have the structure drawn - such structures fall apart by O--S cleavage. Like RCOOH, sulfonic acids protonate on the double bonded O to put charge on S.

24/ I am expecting questions along the lines of "Can't we combine this step with that step and maybe that step to make this happen?" That is the sloppiness of the education of organic chemists, where the teachers shorthand everything to save space or their tired hands and then

25/ students get the idea that the shorthand mechanisms are the real mechanisms. There are two problems with this. The first is that it is mostly not right - there are at least severe limitations on what can happen in single steps - and the second is that it is bad teaching. When

26/students don't have to see intermediates and judge their stability, they start to think that anything at all can happen. It is a very short step from there to drawing methyl and monovalent oxygen cations, or having editors, reviewers, and readers look at them but see nothing.