[Doc Oc]

If you add excess Grignard to an ester you end up with double addition product to get the tertiary alcohol.  In this case, you have a cyclic ester reacting with two Grignards built into the same molecule, which is equivalent to adding excess.

At any rate, yes, your new mechanism looks correct.

[SVXX]

Yeah..thanks a bunch. I won the bet

[johnnyd]

On a side note, are these di-Grignard reagents easy to make? Making it from a dihaloalkane seems impossible. Deprotonation of a diol with 2 equivalents of a Grignard? Do you not just get the bis-alkoxy anion coordinated to a single Mg2⁺? Now I'm curious!

[orgopete]

@SVXX, I agree with the intermediates in your second mechanism. Obviously, the final product is wrong as you removed a carbon. (Your use of curved arrows is also errant. In my class, I would require you to maintain the C-Mg bond as no curved arrows show a change in those electrons nor can you start a curved arrow with an atom, especially Mg²⁺.)

@johnnyd, if the halogens are geminal or vicinal, then you cannot form the Grignard. With longer chains, then you can.

If you deprotonate a diol with a Grignard reagent, with small diols, I expect you would get the chelate as suggested. If the chains are long, and concentrations high, then intermolecular salts are likely to be prevalent. I would use rates of ring formation to be a guide for making this type of prediction.

[SVXX]

@SVXX, I agree with the intermediates in your second mechanism. Obviously, the final product is wrong as you removed a carbon. (Your use of curved arrows is also errant. In my class, I would require you to maintain the C-Mg bond as no curved arrows show a change in those electrons nor can you start a curved arrow with an atom, especially Mg²⁺.)

For someone who has authored an organic chemistry book published by the "Curved Arrow Press", I'd say that was expected aye .
Yes, I noticed that part later. It's actually two five membered rings fused at only one carbon, I made a silly mistake in my excitement there. But I rectified it when I showed it to my mate, forgot to rectify it here though...thanks for pointing it out!

[johnnyd]

Not fully understanding you orgopete.

"if the halogens are geminal or vicinal, then you cannot form the Grignard. With longer chains, then you can." (Sorry not sure how to use the proper quote thing)

I would presume that once the Grignard is formed on one end of the chain and the halogen is present on the other end, either ring closure or intermolecular SN2 reactions (depending on chain length) would proceed? I'm missing something here I reckon. Can you clarify?

Is it a case that Grignard formation (by that I mean insertion of magnesium into a carbon-halogen bond) occurs with a more readily than the unwanted SN2 reactions in which case the di-Grignard could be formed at cold temperature?

[johnnyd]

Anyone?

[orgopete]

Not fully understanding you orgopete.

"if the halogens are geminal or vicinal, then you cannot form the Grignard. With longer chains, then you can."

Because the original statement was general (di-Grignard … seems impossible), I was simply limiting it. If the halogens are on the same carbon (geminal), the Grignard will decompose to form a carbene (I presume). If the halogens are on adjacent carbons (vicinal), then the Grignard will eliminate the second halogen to form an alkene. This reaction is often done with other metals than magnesium, but magnesium will give the same result. If more carbons are present, then dimagnesium salts can form (I didn't look up an references, but I seem to recall seeing this).

I understand how something like this may be difficult to research as no one would try to make a Grignard from dibromomethane, for example, it is then difficult to find what might happen if you did. Incidentally, I believe the Simmons-Smith (Zn + CH₂I₂) may be an example of a (semi) stable zinc equivalent of the intermediate one would get in the magnesium reaction. Because magnesium salts may be considered more ionic, that ionic character leads to the decomposition to magnesium bromide and a carbene. The zinc salts are more stable and thus one can find reactions that may be considered carbenes or carbanions. That is a whole different question though.

[Agathiyar]

Hi SVXX,

It seems that your second scheme is right, though the final product structure is wrong, it should be a spirolactone.