[blaisem]

Hello.

I seem to recall in my undergraduate organic course that the reduction of an ester to an alcohol via LiAlH₄ could be halted at the aldehyde step.

                               1. LiAlH4
                                  (CH₃CH₂)₂O
The reaction was: Ester  Aldehyde
                               2. H₂O

If the second step were left out, then, if my memory serves correctly, the reaction would proceed to the alcohol.  Is this true?

I searched namely for this reaction to double check, and I find primarily the DibAl (Di-isobutyl Aluminum Hydride) reagent is used to achieve isolation of the Aldehyde.

Yet I am still not certain if it is impossible for LiAlH₄, because if I look at the mechanism for LiAlH₄, I see that after the first hydride reduction, the (R-O-AlH₃)^-Li⁺ salt forms.  If water were added, would the salt partition into the aqueous phase, leaving the aldehyde behind without reducing it further?  This is assuming the aldehyde had a reasonable alkyl chain attached to hinder aqueous solubility.

Or maybe adding the water would quench the aluminum hydride.  I also read that protic solvents resulted release of H₂ gas.

Thank you!

[Babcock_Hall]

I am not certain of the answer to your question.  However, I recall seeing a paper in which the authors reduced a phenylester to an aldehyde using a milder reagent than lithium aluminum hydride, namely LiAH(O-tert-Bu)₃.  It might have come out of HC Brown's lab, but I don't have the reference handy.

[blaisem]

Maybe I just imagined it were possible with lithium aluminum hydride then.

The reaction you mentioned might be the dibAl reagent I saw, which basically stops the reaction at the aldehyde in equal molar amounts because the aluminum only has one hydrogen available.  It does have an "electrophilic mechanism" as opposed to a nucleophilic hydride.  It also seems to dimerize with itself.  Here's one draw up of the mechanism:

http://www.masterorganicchemistry.com/2011/08/26/reagent-friday-di-isobutyl-aluminum-hydride-dibal/

[Dan]

It does have an "electrophilic mechanism" as opposed to a nucleophilic hydride. 

This is the key to it's complimentary reactivity compared to LAH. In order to transfer hydride Al in DIBAL must first coordinate to the acyl oxygen. This coordination is faster with more Lewis basic acyl oxygens - this explains why DIBAL reduces amides and esters rapidly in comparison to aldehydes and ketones. So in reduction of an ester with DIBAL, the ester is rapidly reduced to an aldehyde (or hemiacetal) - and because this aldehyde is relatively unreactive to DIBAL it is possible to minimise further (slower) reduction to the alcohol using low temperature.

The problem with nucleophilic reducing agents (like LAH) is that they react fastest with more electrophilic acyl carbons - so aldehydes and ketones react much faster than esters. In practice this means that an ester is relatively slowly reduced to an aldehyde intermediate which is the very rapidly reduced further to the alcohol. This is conceptually similar to the practical difficulty of making a ketone by reaction of a Grignard with an ester - the ketone rapidly reacts again to give the tertiary alcohol.

Similar trends are observed for boron reagents - borane (Lewis acidic) and borohydride (nucleophilic) have similar complimentary reactivity preferences (but are generally less reactive than aluminium reagents).

[Babcock_Hall]

We are talking about two different reagents, but I suspect that both could be made to work under the right conditions.
The paper I was thinking about was probably Weissman and Brown, JOC, 31, 283, 1966, which is cited on p. 73 of H. O. House, Modern Synthetic Reactions, 2nd ed.  This might be a good review:
Reductions by Metal Alkoxyaluminum Hydrides. Part II. Carboxylic Acids and Derivatives, Nitrogen Compounds, and Sulfur Compounds
Jaroslav Málek
Published Online: 15 OCT 2004
DOI: 10.1002/0471264180.or036.03

[blaisem]

Ah thanks for the thorough explanation, Dan.  The correlation with boron analogs was especially interesting.

Does that also mean I was mistaken in the possibility of stopping ester reduction by LiAlH4 at the aldehyde step?

[Babcock_Hall]

Dan, thanks, that was very helpful.  So why does one obtain aldehydes from acid chlorides and phenylesters from Li(O-tert-butyl)₃AlH?  What keeps the aldehyde from reacting further?

[Dan]

LATB is a nucleophilic reducing agent, but is a very weak one. Acyl chlorides are more electrophilic than aldehydes, so it follows that they will react faster than aldehydes with any nucleophilic reducing agent (like LAH). In practice, you can only suppress aldehyde reduction by using a nucleophilic reducing agent that only reacts slowly with aldehydes.

The phenyl ester example you cited earlier is a special case - that will behave somewhere in between an anhydride and an ester (I would expect the electrophilicity to be somewhere in the aldehyde ballpark, and judging by the paper, it is slightly higher).

[orgopete]

It does have an "electrophilic mechanism" as opposed to a nucleophilic hydride. 

This is the key to it's complimentary reactivity compared to LAH. In order to transfer hydride Al in DIBAL must first coordinate to the acyl oxygen. This coordination is faster with more Lewis basic acyl oxygens - this explains why DIBAL reduces amides and esters rapidly in comparison to aldehydes and ketones. So in reduction of an ester with DIBAL, the ester is rapidly reduced to an aldehyde (or hemiacetal) - and because this aldehyde is relatively unreactive to DIBAL it is possible to minimise further (slower) reduction to the alcohol using low temperature.

While I mostly agree with what Dan has posted, I don't agree about an aldehyde's reactivity. I also agree that DIBAL is the reagent to reduce esters to aldehydes and not LiAlH4. The mechanism I had learned for DIBAL reductions is the success of the reductions depends upon the addition of hydride to the ester resulting in a "stable" tetrahedral intermediate. Reversion of the intermediate gives an aldehyde which is readily reduced, just as Grignard reactions. Aldehydes are more reactive. To arrest the reaction at the addition stage, toluene is used as solvent and low temperatures. That helps to stop the elimination of an alkoxide.

If I am wrong, then one should be able to selectively reduce an aldehydo ester to a dialdehyde. I don't recall any way to do that selectively without protection.

[Dan]

While I mostly agree with what Dan has posted, I don't agree about an aldehyde's reactivity. I also agree that DIBAL is the reagent to reduce esters to aldehydes and not LiAlH4. The mechanism I had learned for DIBAL reductions is the success of the reductions depends upon the addition of hydride to the ester resulting in a "stable" tetrahedral intermediate. Reversion of the intermediate gives an aldehyde which is readily reduced, just as Grignard reactions. Aldehydes are more reactive. To arrest the reaction at the addition stage, toluene is used as solvent and low temperatures. That helps to stop the elimination of an alkoxide.

If I am wrong, then one should be able to selectively reduce an aldehydo ester to a dialdehyde. I don't recall any way to do that selectively without protection.

Yes, that is an important point I should have mentioned. In view of your chemoselectivity point I have to agree that this factor (hemiacetal stability) is more important than the difference in reactivity rates of esters and aldehydes with DIBAL. Nevertheless, I maintain that DIBAL reacts faster with more electron rich carbonyls (amides and esters) compared to less electron rich carbonyls (aldehydes and ketones), references to follow.

[Dan]

So, following a poke around I can find no evidence for an answer to the question of relative reactivity of esters and aldehydes with DIBAL.

Reference to DIBAL reacting faster with more Lewis basic compounds is common, but I can find no example of a direct comparison of esters and aldehydes, and as has been pointed out DIBAL does not appear to have ever been used to successfully chemoselectively reduce an ester over an aldehyde (or vice versa - perhaps the rates are comparable).

Based on this, my view is that the stability of the tetrahedral intermediate is the most important factor and that the relative reactivity of an ester vs an aldehyde with DIBAL has not been directly examined. Aldehyde and ketone reductions tend to be run at higher temperature than those of esters - but I do not consider this good evidence since it is not clear whether higher temp is required to reduce an aldehyde or it is simply that low temp is not required because there is no chemoselectivity issue.

I have attached an excerpt from B.M. Trost's Comprehensive Organic Synthesis: Selectivity, Strategy and Efficiency in Modern Organic Chemistry. Reduction, Volume 8, which alludes to the possibility reactivity rate difference factor (but he does not go further than that).

[discodermolide]

My experience with DIBAL reduction of esters run at -70°C usually produces alcohols. If you want it to stop at the aldehyde -90°C or lower is better, but you still get some over reduction.
So it may be easier to reduce to the alcohol and re-oxidise with something like TEMPO/bleach.

[kriggy]

Or maybe adding the water would quench the aluminum hydride.  I also read that protic solvents resulted release of H₂ gas.

Thank you!

I wouldnt add water, unless you want some H₂. If you want to quench the reduction try diethylamine. It should stop the reduction after 1st hydrogen released from LiAlH4. J. Org. Chem, 1987, 52, 5486-5487 I recently tried it in lab, results from MS will come probably tomorow so I will see if it worked. But I had pretty low yield, IMO because of realy old LiAlH4 or short reaction time.

[TwistedConf]

So it may be easier to reduce to the alcohol and re-oxidise with something like TEMPO/bleach.

This has always been my style...  partial reductions seem to work better on paper than in the lab.

[orgopete]

While I mostly agree with what Dan has posted, I don't agree about an aldehyde's reactivity. I also agree that DIBAL is the reagent to reduce esters to aldehydes and not LiAlH4. The mechanism I had learned for DIBAL reductions is the success of the reductions depends upon the addition of hydride to the ester resulting in a "stable" tetrahedral intermediate. Reversion of the intermediate gives an aldehyde which is readily reduced, just as Grignard reactions. Aldehydes are more reactive. To arrest the reaction at the addition stage, toluene is used as solvent and low temperatures. That helps to stop the elimination of an alkoxide.

If I am wrong, then one should be able to selectively reduce an aldehydo ester to a dialdehyde. I don't recall any way to do that selectively without protection.

Yes, that is an important point I should have mentioned. In view of your chemoselectivity point I have to agree that this factor (hemiacetal stability) is more important than the difference in reactivity rates of esters and aldehydes with DIBAL. Nevertheless, I maintain that DIBAL reacts faster with more electron rich carbonyls (amides and esters) compared to less electron rich carbonyls (aldehydes and ketones), references to follow.

Actually, this caught me off guard. I hadn't thought of this possibility.

None the less, selective reductions are commonly run, assuming reasonable stoichiometry is used in DIBAL reductions. If esters react faster than aldehyde, then alcohol should only form due to excess DIBAL. Therefore, alcohol formation should be suppressed by a better stoichiometry. I don't think that is true (though I have not done this reaction). I believe low temperature is chosen to reduce reversion of the intermediate to aldehyde and alcohol forms from its relatively fast reduction to alcohol.