Acetone, CH₃COCH₃, has six hydrogens, and all six sit on the two carbons beside the C=O. Each of them has a pKa of about 20. A hydrogen on the end carbon of propane, which has no C=O, has a pKa of about 50. The only difference is the carbonyl next door.
So far the carbonyl carbon has been an electrophile, the atom a nucleophile attacks. This chapter uses the carbonyl the other way round. It makes the hydrogens next to it acidic enough to remove, and the anion left behind is a carbon nucleophile. That lets one carbonyl compound attack another. Much of the carbon–carbon bond making in living cells works this way.
The alpha carbon and its unusually acidic hydrogens
Look at acetone with every atom drawn. The carbon in the middle, double-bonded to O, is the carbonyl carbon. Each carbon bonded directly to it is an alpha (α) carbon. A hydrogen on an alpha carbon is an alpha hydrogen. The next carbon out, two bonds from the carbonyl carbon, is the beta (β) carbon (for example, C4 of butan-2-one in the worked example below).
Compare the alpha C–H, at pKa 20, with another acidic C–H. A terminal alkyne C–H, from Alkenes & Alkynes, has a pKa of about 25. So a ketone's alpha hydrogen is even more acidic than an alkyne's.
Why: delocalization into the carbonyl
Take one alpha hydrogen off acetone with a base. The two electrons of the C–H bond stay behind on the carbon. That gives a carbanion: a carbon with a lone pair and a negative charge.
This carbanion sits right next to the C=O. Its lone pair can spread into the C=O pi bond, and the pi electrons can move onto the oxygen. So the anion has two resonance contributors. In one, the charge is on the alpha carbon. In the other, the charge is on the oxygen, and a C=C stands where the C–C was. This anion is called an enolate, because it is the anion of an enol (the C=C–OH compound from alkyne hydration, covered again below).
Oxygen is more electronegative than carbon, so it holds a negative charge better. The contributor with the charge on oxygen is the major one.
The enolate still forms its new bonds at carbon. Almost every enolate reaction in this chapter joins the alpha carbon to something. So an enolate keeps most of its charge on oxygen but reacts as a carbon nucleophile. Kinetic & thermodynamic enolates, later in this chapter, looks at why the carbon end usually wins.
Two carbonyls do it again
Pentane-2,4-dione, CH₃COCH₂COCH₃, has a CH₂ between two C=O groups. Those two hydrogens have a pKa of about 9.
Take one of them off and write the contributors. The charge can sit on the central carbon, on the left oxygen, or on the right oxygen. That is three contributors where a simple enolate has two, and two of the three put the charge on an electronegative atom.
The anion is also flat, and its pi system runs the whole way along O–C–C–C–O. So the two carbonyls share the charge instead of competing for it. The pKa drops from 20 to 9: eleven units for one extra C=O.
| Compound | Alpha pKa |
|---|---|
| alkane | ~50 |
| ester | ~25 |
| ketone / aldehyde | ~19–20 |
| β-ketoester | ~11 |
| 1,3-diketone | ~9 |
A 1,3-diketone, at pKa 9, is a little more acidic than phenol, at 10, even though its acidic hydrogen is on carbon. An ordinary alkoxide base removes it completely, with no special reagent. That is why 1,3-dicarbonyl compounds are the everyday carbon nucleophiles of synthesis.
An ester's alpha hydrogen (pKa 25) is less acidic than a ketone's (pKa 20). The ester's second oxygen, the one in the OR group, already donates its lone pair into the C=O. So the C=O has less pull left over for the enolate's charge.
The rule: an alpha carbon is bonded directly to a carbonyl carbon. A hydrogen on the carbonyl carbon itself, the H of an aldehyde's CHO, does not count. Nor does a hydrogen on a carbon one bond further out.
(a) Butan-2-one, CH₃COCH₂CH₃. The carbonyl carbon is C2. So C1 (a CH₃, three H) and C3 (a CH₂, two H) are both alpha. That is five alpha hydrogens in two different sets, which is why this ketone can give two different enolates. C4 is a beta carbon.
(b) 2,2-Dimethylpropanal, (CH₃)₃CCHO. The carbonyl carbon carries an H, but that H sits on the carbonyl carbon, not next to it. The one alpha carbon carries three methyl groups and no hydrogen: zero alpha hydrogens. This aldehyde forms no enolate and no enol, under any conditions.
(c) Ethyl acetate, CH₃CO₂CH₂CH₃. Only the CH₃ on the C=O side is alpha: three H, pKa 25. The OCH₂ carbon is bonded to an oxygen, not to the carbonyl carbon, so its hydrogens are neither alpha nor acidic. Written as a formula, the two groups look alike, so check the bonds.
(d) Pentane-2,4-dione, CH₃COCH₂COCH₃. There are three alpha carbons. The two methyls have the usual pKa of 20. The central CH₂ is alpha to both carbonyls, so its pKa is 9. A base takes a central hydrogen first.
The habit: mark every carbon that touches a C=O carbon. Then count hydrogens on those carbons and nowhere else.
Choosing a base: two regimes
A base can remove a proton whose pKa is below its own pKaH, the pKa of the base's conjugate acid (Acids & Bases). Mix acetone (pKa 20) with hydroxide or ethoxide (pKaH about 16). Either base is weaker than the enolate it would make, so deprotonation is incomplete. A small amount of enolate sits among a large excess of ketone. That is what an aldol reaction needs, because the ketone must be there to act as the electrophile.
Now use LDA, lithium diisopropylamide, from Organolithiums, Cuprates & Cross-Coupling. Its pKaH is about 36, far above 20. Deprotonation is complete and does not reverse. LDA is also bulky, so it does not add to the carbonyl carbon, as a small strong base would.
The base also decides which enolate forms when a ketone has two different alpha carbons, like butan-2-one. That is the subject of Kinetic & thermodynamic enolates, later in this chapter.
The enol tautomer
An enolate shares its charge between carbon and oxygen, so a proton can come back to either atom. A proton on the alpha carbon gives the ketone back. A proton on the oxygen gives the enol: a C=C carrying an OH (“ene” plus “ol”). You met an enol briefly in alkyne hydration, in Alkenes & Alkynes.
Acetone and its enol, prop-1-en-2-ol, have the same atoms, bonded differently. Keto and enol forms are tautomers: constitutional isomers that interconvert quickly because a hydrogen moves. They are not resonance structures, so draw them with ⇌, never ↔.
For acetone, the keto form wins by about 10⁸ to one. A bottle of acetone holds only a few parts per billion of enol. The keto form keeps a C=O, about 178 kcal/mol, where the enol has a C=C, about 147 kcal/mol. That gap outweighs the enol's gain of an O–H, which is a little stronger than the C–H it replaces.
A 1,3-diketone can tip the balance the other way. Pure liquid pentane-2,4-dione is about 80% enol. Its enol C=C is conjugated with the second C=O, and its OH forms a hydrogen bond to that C=O oxygen. Phenol is the extreme case: its enol form is aromatic, so it exists entirely as the enol and not as a ketone.
How the enol forms: base or acid
Pure acetone makes its enol slowly. A trace of base or acid speeds this up. Each route takes two steps, in opposite orders. In base, hydroxide takes an alpha hydrogen first, which gives the enolate. The enolate's oxygen then takes a proton from water, which gives the enol and hands the hydroxide back.
In acid, H₃O⁺ puts a proton on the carbonyl oxygen first. The protonated ketone shares its positive charge between oxygen and the carbonyl carbon. A water molecule then takes an alpha hydrogen, and the C=C forms as it leaves. This gives the enol and hands H₃O⁺ back.
Neither route makes an unstabilized ion. The enolate spreads its negative charge onto oxygen, and the protonated ketone spreads its positive charge onto carbon. The enol is only a trace, but in acid it is the form that reacts at the alpha carbon, as the enolate is in base. The ketone itself is never the nucleophile.
The order follows a rule that holds for every acid- or base-catalyzed mechanism from here on. In acid there is no strong base to pull a proton off, so a proton goes on first; in base there is no strong acid to put one on, so a proton comes off first.
An experiment shows the base route at work. Stir a ketone in D₂O with a trace of NaOD. Every alpha hydrogen, and only the alpha hydrogens, is replaced by deuterium, because each pass through the enolate picks up D instead of H. Acetone exchanges all six. Butan-2-one exchanges five. 2,2-Dimethylpropanal exchanges none. So counting the exchanged hydrogens counts the alpha hydrogens, a question exams ask often.
What enolates do
Once formed, an enolate's alpha carbon acts like any carbon nucleophile. It does SN2 on an alkyl halide (alkylation; SN2 is from Substitution & Elimination). With hydroxide or an alkoxide as the base, this works best on 1,3-dicarbonyls; a simple ketone first needs LDA, because an enolate left beside unreacted ketone attacks that ketone instead. The enolate attacks aldehydes and ketones (the aldol reaction, next section). It attacks esters by acyl substitution (the Claisen condensation, the section after; acyl substitution is from Carboxylic Acids & Derivatives). And it reacts with halogens at the alpha carbon.
One result is worth seeing on its own. (R)-3-Phenylbutan-2-one has its stereocenter on an alpha carbon, and a trace of acid or base racemizes it: it becomes a 50:50 mixture of (R) and (S). Removing the alpha hydrogen turns that carbon into a flat sp² carbon in the enolate, or in the enol under acid. The proton that comes back can arrive from either face, with equal chance.
A stereocenter one carbon further out is safe. In (R)-3-methylcyclohexanone the stereocenter is a beta carbon, so forming the enolate leaves it untouched. And a flat intermediate gives racemization, not inversion. The same chemistry explains why amino acids racemize when a protein is hydrolyzed in hot acid or base, and why a drug with a stereocenter next to a carbonyl can lose its optical purity.
What carries forward
The enolate is the nucleophile for the rest of this chapter. Kinetic versus thermodynamic control returns wherever two products compete under different conditions. And because a 1,3-dicarbonyl is acidic enough for an ordinary base, the acetoacetic ester and malonic ester syntheses work. These classic routes to substituted ketones and carboxylic acids are the subject of the second-to-last section of this chapter.