[qw098]

Hi Guys,

I've been thinking... so I have a ketone... and if I react a ketone with a strong non-nucleophilic base like LDA I form the enolate.

But lets say I use NaOH... I again form the enolate. My question is... can't the NaOH also act as a nucleophile and add to the carbonyl? I know I can hydrolyze ketones with H2O to form a diol... why can't I simply react NaOH with a ketone and form a diol?

This is just me thinking and I tried googling the answer... but no answer! I wanted to know your guys' opinion then I will ask my professor tomorrow

Thanks!

[Doc Oc]

I think you're confused about what you can/can't make with a ketone.

Yes, NaOH can act as a nucleophile to a ketone.  You will end up with a geminal diol.  These rapidly return back to ketones.  If you look up the mechanism for the deprotection of a ketal, you will see how this works (also, if you look up how a ketal is made, you can work backwards and see why adding HO- doesn't do anything).

Ketones are not hydrolyzed to diols by water, I'm not sure where you saw/heard that.

[Dan]

My question is... can't the NaOH also act as a nucleophile and add to the carbonyl?

Of course. That is ketone hydration, and it is reversible. This is like the opening step of the Cannizzaro reaction (which involves an aldehyde).

I know I can hydrolyze ketones with H2O to form a diol... why can't I simply react NaOH with a ketone and form a diol?

Do you really mean hydrolyse? Hydration is possible:

RCOR + H₂O RC(OH)₂R

but it's reversible and the equilibrium normally lies far to the left.

Can you indicate which diol you are talking about?

[qw098]

Ketones are not hydrolyzed to diols by water, I'm not sure where you saw/heard that.

Sorry, I meant hydrated to a diol. My apologies.

Of course. That is ketone hydration, and it is reversible. This is like the opening step of the Cannizzaro reaction (which involves an aldehyde).

Ok, so NaOH can act as a nucleophile but the equilibrium lies so much to the left that the NaOH ends up acting as a base to form the corresponding enolate!?

However, reacting a ketone with water will yield the diol we want quite easily because the equilibrium lies far to the right? Right?!

[Dan]

Ok, so NaOH can act as a nucleophile but the equilibrium lies so much to the left that the NaOH ends up acting as a base to form the corresponding enolate!?

I would argue that it is not the position of the equilibrium for intermediate formation, but the productivity of the intermediate that is important. Remember, the position of equilibrium for the deprotonation of a ketone (pKa ~ 20) with NaOH (pKa of water ~16) lies far to the left as well.

You cannot normally isolate hydrated ketones, and you cannot normally isolate enolates by treating ketones with NaOH (you can if you use stronger bases though).

Point is, the enolate intermediate reacts to form more stable products, whereas a hydrated ketone is usually a dead end intermediate.

I gave the example of a Cannizzaro for an important reaction in which addition of hydroxide to an aldehyde is productive. Also, it occurs to me now that the reverse-Claisen condensation (which is a ketone [specifically a b-ketoester] hydrolysis reaction) is an example in which this type of addition of hydroxide to a ketone does lead to a thermodynamic minimum.

However, reacting a ketone with water will yield the diol we want quite easily because the equilibrium lies far to the right?

No, it still lies to the left.

[qw098]

Great stuff Dan!

Very interesting stuff. I never knew I couldn't isolate hydrated ketones.

Point is, the enolate intermediate reacts to form more stable products, whereas a hydrated ketone is usually a dead end intermediate.

Great point, makes a lot of sense! I feel like my understanding of reactions in organic chemistry have greatly improved. The moral of the story is... "noone likes dead-end intermediates" .

Thanks again Dan!

[Cryolite]

You could, in some special cases where the hydrate is very stable, such as chloral.

[Dan]

Cyclopropanone is also more stable as the hydrate far as I remember...