3/12/2007

Excimer doesn't know everything

I hate to burst anyone's bubble, but, contrary to what my mom thinks, I don't know all organic chemistry. I can usually bullshit my way out of most organic problems, but this one continues to stump me. My undergrads do this Grignard reaction where they react phenylmagnesium bromide that they make in situ, then add ethyl benzoate to make tritylcarbinol (which is what old people call triphenylmethanol). They get, as one minor byproduct, benzhydrol.



I have some guesses, and I've searched the literature for possible explanations- the best I can come up with is this paper (Noller, C.N. JACS 1931, 53, 635-643) which explains reduction in the presence of an alkyl Grignard, but I don't see that happening here. I suspect it has something to do with the radical nature of the Grignard reagent, but I have no concrete evidence. And I'm really too lazy to dive further into the literature to find a reasonable explanation.

Anyone know? Show your work. Include references. kthx.

(I think this whole thing goes to show, as was laid bare in the comments of this latest post from Chembark, that the organic chemistry we learned as sophomores is waaaaaaay more complex than how it was taught to us. It's hard to garner an appreciation of the complexity of organic chemistry when you're learning it for the first time and not be turned off, or at least frusturated, by it. I consider it a real challenge of teaching chemistry in general.)

16 comments:

Liquidcarbon said...

Single electron transfer (should I say one-electron reduction?). I wonder if the solution was colored.

Analogous to this

Excimer said...

The solution is highly colored. The PhMgBr in situ sort of a gross brown upon formation and it turns really pretty reddish pink immediately upon addition of ethyl benzoate, which disappears gradually as the reaction progresses.

The pinacol-type single electron reduction was my prof's guess but that doesn't explain where the hydrogens come from.

Liquidcarbon said...

From ethoxide. Probably, some Mg(OH)2 is around

milkshake said...

Are you sure it is benzhydrol - and not binaphtyl for example? If you make Grignard from bromobenzene there is always quite a bit of biphenyl, we have had problems to remove it from some unpolar products so we switched to PheMgCl in THF which does not have the problem.

The only explanation I could come up for formation of benzhydrol would be if you had some unreacted Mg around and formed benzophenon-ketyl anion radical Mg salt (analogous to sodium ketyl, when benzophenon is used as indicator in Na/THF stills)

milkshake said...

biphenyl, sorry

Excimer said...

milkshake- biphenyl is definitely another byproduct, but they also make this one.

evgeny said...

I agree with milkshake. Since the Grignard is made in situ, you've got a whole lot of magnesium floating in there. From literature, this is a common reaction back in the 30s and 20s when people treated benzophenone with various alkyl Grignards (that's all the literature search work I did for this one, sorry...). Probably an electron transfer from magnesium. It happens readily enough with sodium. The radical salt can dimerize, but then abstracts four hydrogens from water upon quenching to give two alcohols (breaking apart on quenching) and two Mg(OH)2. We do use it in stills, so that explanation makes sense...

Biphenyl is from reductive elimination from Mg(Ph)2, which is formed in the Grignard equilibrium. Although, I don't think anyone studied this reductive elimination in much detail (i.e. it might happen only when you add water). It would be an interesting thing to look into actually. It is possible to separate the equilibrium mixture and get pure dialkyl or diaryl magnesium reagent (or magnesium dibromide for that matter).

There would be a way to test my theory. You'll have to use premade PhMgBr, or just carefully use a cannula to transfer it, making sure no small magnesium pieces also get in there. There should be a lot less of this byproduct then.

Excimer said...

evgeny- It's called the Schlenk equilibrium (Schlenk discovered the equilibrium), and I've heard conflicting reports on how biphenyl is formed. Your explanation- reductive elimination of PhMgPh- makes the most sense to me, but I've heard other plausible reasons (that I can't remember off the top of my head, but they were sound). Then again, reductive elimination of two ligands that are 180° apart from one another suggests an intermediate is involved. I dunno for sure. I wonder if anyone does?

I agree, though- single electron transfer makes the most sense given what's in the flask at the time (excess Mg, especially). But the proton abstraction still doesn't sit well with me.

Mitch said...

You can always ask questions at Chemical Forums. :P

Mitch

Excimer said...

I COULD post at chemicalforums-

Or I could continue my boycott of all forums (nothing against it, aside from my disdain of all forums) AND actually post something on this blog, which I ever-so-not-as-frequently-as-I'd-like do.

evgeny said...

Thanks for reminding me of the name! It just popped out of my head and I couldn't remember for the life of me. Unless I'm wrong, not much is known about how 'really' the Schlenk equilibrium works and how the halogen and the aryl groups are transferred between metal centers. If you can figure it out though, it's a straight path to Andjewandte whether you did it by experimental or computational means.

I'm putting my money down on sigma bond metathesis. Plus, with this mechanism, your products are not at that awkward 180 degree angle right after the reaction takes place. This would mean however, that reductive elimination to produce the starting aryl halide also takes place. However, the magnesium metal produced would be free in solution and be highly reactive, so the Grignard would be regenerated eventually. Plus, reductive elimination might not take place at low fridge temperatures at which most Grignards are stored. In an undergrad lab where they are made in situ however... it's a bit of a different story.

Hmmm.... this gives me an idea for a weekend computational project. Mind you, I've had a bit of wine, so maybe I'll think it's all a bit of tosh in the morning (that, or it's been done before).

milkshake said...

Wine is for ballet girls, real men have 250 grams of Samogon for breakfest. (It fills one's eyes with tears but clears the head from the last nights hungover).

bchem said...

Could you be seeing the Mg equivalent of the Oppenauer Oxidation / Meirwein Pondorf Varley reduction?

LC said...

Oh, Milkshake. Don't remind me.

evgeny said...

Well milkshake, when I come over to Palm Beach country to visit, you can take all the 200 proof ethanol, DI water, and cranberry juice you want and mix them together to celebrate my visit. Until then, I'm sticking with the sophisticated stuff: Wine. If it's good enough for the French, it's good enough for me.

I think that the reason "real men" had samogon for breakfast was because alcoholic drinks were outlawed in the 80s, so you had to make your own. It's easier if you're a chemist.

Anonymous said...

The reductive elimination to form biphenyl makes sense. The actual diaryl magnesiums (magnesia?) aren't linear. They're pretty highly aggregated, so they'd be on the order of 100 degrees apart. Close enough to reductively eliminate.