Wittig Reaction

For the discussion of the sciences. Physics problems, chemistry equations, biology weirdness, it all goes here.

Moderators: gmalivuk, Moderators General, Prelates

User avatar
BlackSails
Posts: 5315
Joined: Thu Dec 20, 2007 5:48 am UTC

Wittig Reaction

Postby BlackSails » Sun May 11, 2008 12:09 am UTC

In molecules with multiple carbonyl groups, is it possible to determine (non-experimentally) which (if the reaction is selective at all) carbonyl will be drawn into the oxaphosphatane, and replaced with the alkyl group? Ketones vs Aldehydes, Esters, acids, etc?

I would guess that acids > esters > aldehydes > ketones. Esters are more sterically hindered than acids. They both have more electrophilic carbonyl carbons than ketones and aldehydes. Aldehydes are less sterically hindered than ketones. Esters and (conjugate bases of) acids however, are very stable due to resonance, which might overcome the effects of that electrophilic carbon.

So, is it possible to selectively alkylate one of several carbonyl groups in a molecule (or equivalently, a mixture of molecules)?

PS. Not homework, just a question that came up during a practice synthesis question.

psyck0
Posts: 1651
Joined: Wed Mar 19, 2008 5:58 pm UTC
Contact:

Re: Wittig Reaction

Postby psyck0 » Sun May 11, 2008 6:59 am UTC

The general reactivity order of carboxyl compounds is acid halides> aldehydes > ketones > acids > esters > amides, and I would expect this to follow the same chain. Amides are very difficult to react, while acid halides will do everything in sight. Generally, the fact that acids, esters and amides are more electrophilic is more than counterbalanced by their high stabilities from conjugation of the double bond. I don't think that sterics is a factor. I'm also not 100% certain that this would work with an acid halide, because the halide tends to leave.

However, I am still not certain it is possible to selectively alkylate only one group. Could you protect the other(s) in some way? What is the original question?

Edit: Wikipedia confirmed my suspicions. The Wittig reaction only works with aldehydes and ketones.

God, I love organic chemistry.

User avatar
BlackSails
Posts: 5315
Joined: Thu Dec 20, 2007 5:48 am UTC

Re: Wittig Reaction

Postby BlackSails » Sun May 11, 2008 11:50 am UTC

psyck0 wrote:The general reactivity order of carboxyl compounds is acid halides> aldehydes > ketones > acids > esters > amides, and I would expect this to follow the same chain. Amides are very difficult to react, while acid halides will do everything in sight.


Amides have their godawful rearragement reactions though.

Generally, the fact that acids, esters and amides are more electrophilic is more than counterbalanced by their high stabilities from conjugation of the double bond. I don't think that sterics is a factor. I'm also not 100% certain that this would work with an acid halide, because the halide tends to leave.


Well, sterics are why aldehydes are more reactive than ketone. I assume the same would apply for ester/acid.

However, I am still not certain it is possible to selectively alkylate only one group. Could you protect the other(s) in some way?


The only way I know to protect carbonyls is acetal formation, which isnt selective at all, and only works in base. I suppose you could protect them with a wittig reaction, and then recover the carbonyl by ozonolysis w/reductive workup, but you better not have any other double bonds in your molecule. But again, not selective.

God, I love organic chemistry.


I hated it, then loved it (aromatic reactions are the best), and how that im studying rearragements and pericyclic reactions, I hate it again.

Speaking of hate, what the hell is up with fullvene synthesis? Cyclopentadienyl anion adds to carbonyl. Carbonyl is protonated. That all makes sense. Cyclopentadienyl anion is a lewis base, the carbonyl carbon is a lewis acid. But then you deprotonate the ring again (at the tertiary site), and kick out hydroxide(!) as a leaving group to form a double bond? No other reaction in the book kicks off hydroxide.

psyck0
Posts: 1651
Joined: Wed Mar 19, 2008 5:58 pm UTC
Contact:

Re: Wittig Reaction

Postby psyck0 » Sun May 11, 2008 6:55 pm UTC

Well, sterics are why aldehydes are more reactive than ketone. I assume the same would apply for ester/acid.


But the difference between an aldehyde and a ketone is large- a tiny H atom vs a much larger CX3 atom attached to the carbonyl carbon. The difference between an acid and an ester is much smaller, because it is one atom removed from the carbonyl group (unless, of course, you have a very bulky ester). I think that I may have been wrong, and that acids/esters are at the same level of reactivity, but I don't have my notes to confirm.

Acetals and ketals only work with aldehydes and ketones, respectively, don't they? Use glycerol to make a 5-membered ring ketal/acetal. I think you can do it in acid, can't you? I really wish I had my notes, but Wikipedia seems to agree with me.

Did the "wittig reaction only works on aldehydes and ketones" solve your problem? If not, what's the original question?

I don't remember doing fulvene synthesis, but from the sounds of the reaction, hydroxide is the only sensible leaving group. It's certainly more stable than a carbanion, even an aromatic one. I also remember one reaction we covered kicking off hydroxide, but I don't remember which one. That was for the same reason- hydroxide was the only possible leaving group, and the product was very stable.

I don't see why you dislike rearrangements. They're pretty simple as long as you're careful. Pericyclic reactions are confusing, though. Trying to figure out which bit ends up where- eugh.

opsomath
Posts: 89
Joined: Wed Nov 01, 2006 1:58 pm UTC

Re: Wittig Reaction

Postby opsomath » Tue May 13, 2008 1:32 pm UTC

Sterics have relatively little to do with carbonyl reactions; what matters more is the relative electrophilicity of the carbonyls. Aldehyde tops this list by a lot; I would expect (offhand guess) 90% selectivity for an aldehyde over a ketone in a Wittig.

Hydroxide is really not such a bad leaving group; they tell you it is in beginning orgo, but that's only relative to stuff like triflate or chloride or something. In the case of cyclopentadienyl, the two are almost the same basicity thanks to aromaticity; what drives the reaction is that the product (fulvene) is more stable, so in the case where the right series of reactions do occur with a single pair of molecules, it'll probably stay there.

I assume you're doing this reaction in an organic solvent? Sodium cyclopentadienyl plus, say, formaldehyde? The driving force here is the breaking of the C=O double bond, which is relatively weak, and replacing it with the more favorable bonding situation in fulvene. That, and the precipitation of sodium hydroxide; once the forward reaction takes place, it can't easily reverse.

psyck0
Posts: 1651
Joined: Wed Mar 19, 2008 5:58 pm UTC
Contact:

Re: Wittig Reaction

Postby psyck0 » Wed May 14, 2008 1:06 am UTC

What are you talking about, weak C=O bond? CO2 is least reactive form of carbon (well, perhaps short of diamonds) for a reason. And I think that hydroxide IS a bad leaving group, just not so bad relative to other things that generally wouldn't be considered leaving groups at all.

I was surprised to learn that the pKa of cyclopentadiene is 16, though. That's really low!

opsomath
Posts: 89
Joined: Wed Nov 01, 2006 1:58 pm UTC

Re: Wittig Reaction

Postby opsomath » Wed May 14, 2008 11:59 am UTC

Sorry, phrased that badly. I refer to the pi bond in the double bond. In general, reactions that replace a double bond (pi plus sigma) with two single bonds have that as a driving force, so a double bond is weaker than two single bonds.

In this case, though, I suspect the carbon-carbon double bond is more favorable since it allows for an aromatic resonance form in the ring; to do that with the oxygen there, you'd have to put a positive charge on oxygen, which is bad.


Return to “Science”

Who is online

Users browsing this forum: No registered users and 11 guests