building on: arriving at - chtf.stuba.skszolcsanyi/education/files/organicka chemia ii... · meo me...

Chapters 26–29 continue the theme of synthesis that started with Chapter 24 and will end withChapter 30. This group of four chapters introduces the main C–C bond-forming reactions of enolsand enolates. We develop the chemistry of Chapter 21 with a discussion of enols and enolates attack-ing to alkylating agents (Chapter 26), aldehydes and ketones (Chapter 27), acylating agents (Chapter28), and electrophilic alkenes (Chapter 29).

Carbonyl groups show diverse reactivityIn earlier chapters we discussed the two types of reactivity displayed by the carbonyl group. We firstdescribed reactions that involve nucleophilic attack on the carbon of the carbonyl, and in Chapter 9we showed you that these are among the best ways of making new C–C bonds. In this chapter weshall again be making new C–C bonds, but using electrophilic attack on carbonyl compounds: inother words, the carbonyl compoundwill be reacting as the nucleophile inthe reaction. We introduced thenucleophilic forms of carbonyl com-pounds—enols, and enolates—inChapter 21. There you saw themreacting with heteroatomic elec-trophiles, but they will also react wellwith carbon electrophiles providedthe reaction is thoughtfully devised.Much of this chapter will concern thatphrase, ‘thoughtfully devised’.

Thought is needed to ensure that the carbonyl compound exhibits the right sort of reactivity.In particular, the carbonyl compound must not act as an electrophile when it is intended to be anucleophile. If it does, it may react with itself to give some sort of dimer—or even a polymer—ratherthan neatly attacking the desired electrophile. This chapter is devoted to ways of avoiding this: inChapter 27 we shall talk about how to promote and control the dimerization, known as the aldolreaction.

26Alkylation of enolates

Connections

Building on:

• Enols and enolates ch21

• Electrophilic addition to alkenes ch20

• Nucleophilic substitution reactionsch17

Arriving at:

• How to make new C–C bonds usingcarbonyl compounds as nucleophiles

• How to prevent carbonyl compoundsreacting with themselves

Looking forward to:

• Forming C–C bonds by reactingnucleophilic enolates withelectrophilic carbonyl compoundsch27

• Forming C–C bonds by reactingnucleophilic enolates withelectrophilic carboxylic acidderivatives ch28

• Forming C–C bonds by reactingnucleophilic enolates withelectrophilic alkenes ch29

• Retrosynthetic analysis ch30

R

O

H

R

NuO

R

NuOH

Nu

carbonyl compound acts as an electrophile

R

O

H

R

O

R

O

EEB

enolate acts as a nucleophile

enolate

Fortunately, over the last three decades lots of thought has already gone into the problem of con-trolling the reactions of enolates with carbon electrophiles. This means that there are many excellentsolutions to the problem: our task in this chapter is to help you understand which to use, and whento use them, in order to design useful reactions.

Some important considerations that affect all alkylationsThese reactions consist of two steps. The first is the formation of a stabilized anion—usually (but notalways) an enolate—by deprotonation with base. The second is a substitution reaction: attack of thenucleophilic anion on an electrophilic alkyl halide. All the factors controlling SN1 and SN2 reactions,which we discussed at length in Chapter 17, are applicable here.

In each case, we shall take one of two approaches to the choice of base.

• A strong base can be chosen to deprotonate the starting material completely. There is completeconversion of the starting material to the anion before addition of the electrophile, which is addedin a subsequent step

• Alternatively, a weaker base may be used in the presence of the electrophile. The weaker base willnot deprotonate the starting material completely: only a small amount of anion will be formed,but that small amount will react with the electrophile. More anion is formed as alkylation uses itup

The second approach is easier practically (just mix the starting material, base, and electrophile),but works only if the base and the electrophile are compatible and don’t react together. With the firstapproach, which is practically more demanding, the electrophile and base never meet each other, sotheir compatibility is not a concern. We shall start with some compounds that avoid the problem ofcompeting aldol reactions completely, because they are not electrophilic enough to react with theirown nucleophilic derivatives.

Nitriles and nitroalkanes can be alkylatedProblems that arise from the electrophilicity of the carbonyl group can be avoided by replacing C=Oby functional groups that are much less electrophilic but are still able to stabilize an adjacent anion.We shall consider two examples, both of which you met in Chapter 21.

Alkylation of nitrilesFirstly, the nitrile group, which mirrors the carbonyl group in general reactivity but is much less eas-ily attacked by nucleophiles (N is less electronegative than O).

664 26 . Alkylation of enolates

R

O

R

O

R

O

ROH

R

O R OH R OHrepeat enolization and attack

aim to avoid an unwanted dimerization: the aldol reaction

nucleophilic enolate

electrophilic carbonyl

unwanted "aldol product"

unwanted polymern

R1

O

H

R1

O

R1

O

R2R2 XB

step 1: formation of enolate anion step 2: alkylation (SN2 reaction with alkyl halide)

strong base(complete formation

of anion)

weak base(anion in equilibrium with

starting material) alkyl halide

SN2

LYou met nitrile hydrolysis and additionreactions, for example, in Chapter 12.

The anion formed by deprotonating a nitrile using strong base will not react with other moleculesof nitrile but will react very efficiently with alkyl halides. The slim, linear structure of the anionsmakes them good nucleophiles for SN2 reactions.

The nitrile does not have to be deprotonated completely for alkylation: with sodium hydroxide onlya small amount of anion is formed. In the example below, such an anion reacts with propyl bromide togive 2-phenylpentanenitrile.

This reaction is carried out in a two-phase mixture (water + an immiscible organic solvent) toprevent the hydroxide and propyl bromide merely reacting together in an SN2 reaction to givepropanol. The hydroxide stays in the aqueous layer, and the other reagents stay in the organic layer.A tetraalkylammonium chloride (benzyltriethylammonium chloride BnEt3N+Cl–) is needed as aphase transfer catalyst to allow sufficient hydroxide to enter the organic layer to deprotonate thenitrile.

Nitrile-stabilized anions are so nucleophilic that they will react with alkyl halides rather well evenwhen a crowded quaternary centre (a carbon bearing no H atoms) is being formed. In this examplethe strong base, sodium hydride, was used to deprotonate the branched nitrile completely and benzylchloride was the electrophile. The greater reactivity of benzylic electrophiles compensates for thepoorer leaving group. In DMF, the anion is particularly reactive because it is not solvated (DMF sol-vates only the Na+ cation).

The compatibility of sodium hydride with electrophiles means that, by adding two equivalentsof base, alkylation can be encouraged to occur more than once. This dimethylated acid was requiredin the synthesis of a potential drug, and it was made in two steps from a nitrile. Double alkylationwith two equivalents of NaH in the presence of excess methyl iodide gave the methylated nitrilewhich was hydrolysed to the acid. The monoalkylated product is not isolated—it goes on directly tobe deprotonated and react with a second molecule of MeI.

Nitriles and nitroalkanes can be alkylated 665

R1 R2R1R1 HR1

C

R2 X

N

C

N

C

N

C

N

Bnitrile ‘enolate’ anionnitrile

alkyl halide

new C–C bond

alkylationdeprotonation

CN

CNBr

NaOHBnEt3N Cl

35 °C

84% yield

LYou met phase transfer catalysis inChapter 23, p. 000.

P

Remember our discussion aboutthe lack of nucleophilicity ofhydride (H–) in Chapter 6? Here ishydride acting as a base even inthe presence of the electrophile:there was no need to do thisreaction in two steps because thebase and electrophile cannotreact together.

N Ph

OCN

N Ph

OCNPh

Cl

Ph

NaH, DMF, –10 °Cquaternary carbon atom marked

MeOCN MeO

CN

Me Me

MeOCO2H

Me MeNaH × 2

MeI

H2SO4

With two nitrile groups, the delocalized anion is so stable that even a weak, neutral amine (tri-ethylamine) is sufficiently basic to deprotonate the starting material. Here double alkylation againtakes place: note that the electrophile is good at SN2, and the solvent is dipolar and aprotic (DMSOand DMF have similar properties). The doubly alkylated quaternary product was formed in 100%yield.

If the electrophile and the nitrile are in the same molecule and the spacing between them is appro-priate, then intramolecular alkylation will lead to cyclization to form rings that can have anythingfrom three to six members. The preparation of a cyclopropane is shown using sodium hydroxide asthe base and chloride as a leaving group. With an intramolecular alkylation, the base and the elec-trophile have to be present together, but the cyclization is so fast that competing SN2 with HO– is nota problem.

Alkylation of nitroalkanesThe powerful electron-withdrawing nature of the nitro group means that deprotonation is possibleeven with very mild bases (the pKa of MeNO2 is 10). The anions react with carbon electrophiles anda wide variety of nitro-containing products can be produced. The anions are not, of course, enolates,but replacing the nitrogen with a carbon should help you to recognize the close similarity of thesealkylations with the enolate alkylations described later.

Surprisingly few simple nitroalkanes are commercially available but more complex examples canbe prepared readily by alkylation of the anions derived from nitromethane, nitroethane, and 2-nitro-propane. Deprotonation of nitroalkanes with butyllithium followed by the addition of alkyl halidesgives the alkylated nitroalkanes in good yield. Some examples of this general method are shownbelow. These reactions really do have to be done in two steps: BuLi is not compatible with alkylhalides!


MeO

H HH

N

MeO•

N

Me I

MeO

H Me

N

H

first alkylation

MeO

H MeH

N

MeO•

N

Me I

MeO

Me Me

N

Me

second alkylation

P

Multiple alkylation is not alwaysdesirable, and one of the side-reactions in alkylations that areintended to go only once is theformation of doubly, or in specialcases triply, alkylated products.These arise when the firstalkylation product still has acidicprotons and can be deprotonatedto form another anion. This may inturn react further. Clearly, this ismore likely to be a problem if thebase is present in excess and canusually be restricted by using onlyone equivalent of theelectrophile.

Cl

PhNC CN

Ph Ph

CNNCEt3N, DMSO100% yield

+

CNCl CNCl

HHHO

N•Cl

N

NaOH0–100 °C

NOO O

nitro-stabilized anion —compare enolate

LNitro-stabilized anions also undergoadditions to aldehydes, ketones, andelectrophilic alkenes: these reactionsappear in Chapters 27 and 29.

N

R1

OON

R1 H

OON

R1

OO

R2 X

N

R1 R2

OO

NR3

nitroalkane nitro-stabilized anion

weak amine base

Nitroalkanes can be alkylated in a single step with hydroxide as a base: phase transfer conditionskeep the HO– and the electrophile apart, preventing alcohol formation. This compound formsdespite its quaternary carbon atom.

Cyclic nitroalkanes can be prepared by intramolecular alkylation provided that the ring size isappropriate (3–7 members). Now there really is no alternative: the base and electrophile mustcohabit in the reaction mixture, so aweaker base such as potassium car-bonate must be used—amines are nogood here because they undergo sub-stitution reactions with the halide.

Choice of electrophile for alkylationEnolate alkylations are SN2 reactions (polar solvents, good charged nucleophile) so the electro-phile needs to be SN2-reactive if the alkylation is to succeed: primary and benzylic alkyl halidesare among the best alkylating agents. More branched halides tend to prefer to undergo unwantedE2 elimination reactions (Chapter 19), because the anions themselves are rather basic. As a result, ter-tiary halides are useless for enolate alkylation. We shall see a way round this problem later in the chapter.

Lithium enolates of carbonyl compoundsThe problem of self-condensation of carbonyl compounds (that is, enolate reacting with unenolizedcarbonyl) under basic conditions does not exist if there is absolutely no unenolized carbonyl com-pound present. One way to achieve this is to use a base sufficiently strong (pKa at least 3 or 4 unitshigher than pKa of the carbonyl compound) to ensure that all of the starting carbonyl is convertedinto the corresponding enolate. This will work only if the resulting enolate is sufficiently stable tosurvive until the alkylation is complete. As you saw in Chapter 21, lithium enolates are stable, and areamong the best enolate equivalents for use in alkylation reactions.

Lithium enolates of carbonyl compounds 667

MeNO2

NO2

I

1. BuLi, THF, HMPA

2.

NO2 NO2

I

NO2 NO21. BuLi, THF, HMPA

2. PhCH2Br

PHexamethylphosphoramide(HMPA)

HMPA has the structure shownbelow, and the basic oxygen atomcoordinates to lithium extremelypowerfully. The cation is solvated,leaving the anion unsolvated andmore reactive. HMPA is known tocause cancer, and should not beconfused with its less commoncousin HMPT (hexamethyl-phosphorous triamide).

O

PMe2N

Me2NNMe2

HMPA

PMe2N

Me2NNMe2

HMPTO2NNO2

O2NNO2

Cl

Bu4NOH

H2O, benzene+

NO2

NO2Br

K2CO3

benzene

H3C X X X X

R X

R1

R2

X

R1

R2

R3

methyl allyl benzyl primary alkyls secondary alkyls tertiary alkyls

alkylate very well alkylate well alkylate slowly do not alkylate

The best base for making lithium enolates is usually LDA, made from diisopropylamine(i-Pr2NH) and BuLi. LDA will deprotonate virtually all ketones and esters that have an acidic protonto form the corresponding lithium enolates rapidly, completely, and irreversibly even at the low tem-peratures (about –78°C) required for some of these reactive species to survive.

Deprotonation occurs through acyclic mechanism illustrated belowfor ketones and esters. The basicnitrogen anion removes the protonas the lithium is delivered to theforming oxyanion.

Alkylations of lithium enolatesThe reaction of these lithium enolates with alkyl halides is one of the most important C–C bond-forming reactions in chemistry. Alkylation of lithium enolates works with both acyclic and cyclicketones as well as with acyclic and cyclic esters (lactones). The general mechanism is shown below.

Typical experimental conditions for reactions of kinetic enolates involve formation of the enolate atvery low temperature (–78°C) in THF. Remember, the strong base LDA is used to avoid self-conden-sation of the carbonyl compound but, while the enolate is forming, there is always a chance thatself-condensation will occur. The lower the temperature, the slower the self-condensation reaction,and the fewer by-products there are. Once enolate formation is complete, the electrophile isadded (still at –78°C: the lithium enolates may not be stable at higher temperatures). The reactionmixture is then usually allowed to warm up to room temperature to speed up the rate of the SN2alkylation.


LLDA is described on p. 000.

N

H

N

Li

BuLi, THF, 0 °C

+ BuH

a reminder: how to make LDA

LDAbutane

diisopropylamine

OR2

H

O

R1

OR2

OLi

R1

LiN

i -Pr

i -Pr

H+ i -Pr2NH

lithium enolate: two geometries possible

deprotonation

–78 °CTHF

deprotonation of an ester

H

O

R1

OLi

R1

LiN

i -Pr

i -Pr

R2

HH H

R2

+ i -Pr2NH

lithium enolate: two geometries possible

deprotonation

–78 °CTHF

deprotonation of a ketone

if R1 ≠ R2, removal of the green protons gives a different enolate

P

Enolates are a type of alkene, andthere are two possiblegeometries of the enolate of anester. The importance of enolategeometry is discussed in Chapter34 and will not concern us here.More important is the question ofregioselectivity whenunsymmetrical ketones aredeprotonated. We shall discussthis aspect later in the chapter.

Variations on a theme

LDA came into general use in the 1970s, and you maymeet more modern variants derived from butyllithium andisopropylcyclohexylamine (lithium isopropylcyclo-hexylamide, LICA) or 2,2,6,6-tetramethylpiperidine

(lithium tetramethylpiperidide, LTMP) orhexamethyldisilazane (lithium hexamethyldisilazide,LHMDS), which are even more hindered and are even lessnucleophilic as a result.

N

Li

LICA

NSi Si

Me

Me

Me

Me

Me MeLi

LHMDS

N

Li

LTMP or LiTMP

OR2

O

R1

LiI

Me

alkylation of an ester enolate

R2

O

R1

LiI

Me

alkylation of a ketone enolate

OR2

Me

O

R1 + LiI

R2

Me

O

R1 + LiI

Alkylation of ketonesPrecisely this sequence was used to methylate the ketone below with LDA acting as base followed bymethyl iodide as electrophile.

In Chapter 17 you saw epoxides acting as electrophiles in SN2 reactions. They can be used to alky-late enolates providing epoxide opening is assisted by coordination to a Lewis acidic metal ion: inthis case the lanthanide yttrium(III). The new C–C bond in the product is coloured black. Note thatthe ketone starting material is unsymmetrical, but has protons only to one side of the carbonylgroup, so there is no question over which enolate will form. The base is one of the LDA variants weshowed you on p. 000—LHMDS.

Alkylation of estersIn Chapter 28 you will meet the reaction of an ester with its own enolate: the Claisen condensation.This reaction can be an irritating side-reaction in the chemistry of lithium ester enolates when alky-lation is desired, and again it can be avoided only if the ester is converted entirely to its enolate underconditions where the Claisen condensation is slow. A good way of stopping this happening is to addthe ester to the solution of LDA (and not the LDA to the ester) so that there is never excess ester for theenolate to react with.

Alkylations of lithium enolates 669

O

OEt

O

OEt

Me1. LDATHF, –78 °C

2. MeI–78 °C to 0 °C

93% yield

O O

HOO

OO

Y3 LiO

H

Li(Me3Si)2N

1. LiN(SiMe3)2

2. Lewis acidic Y(III) salt

99% yield

Lewis acid coordinates to O and assists ring opening

H2O

new C–C bond

Sodium and potassium also give reactive enolates

Their stability at low temperature means that lithiumenolates are usually preferred, but sodium and potassiumenolates can also be formed by abstraction of a proton bystrong bases. The increased separation of the metalcation from the enolate anion with the larger alkali metalsleads to more reactive but less stable enolates. Typicalvery strong Na and K bases include the hydrides (NaH, KH)or amide anions derived from ammonia (NaNH2, KNH2) or

hexamethyldisilazane (NaHMDS, KHMDS). The instabilityof the enolates means that they are usually made andreacted in a single step, so the base and electrophileneed to be compatible. Here are two examples ofcyclohexanone alkylation: the high reactivity of thepotassium enolate is demonstrated by the efficienttetramethylation with excess potassium hydride andmethyl iodide.

O

Br

O OKH excessMeI excess

81% yield

NaNH2Et2O

Another successful tactic is to make the group R as large as possible to discourage attack at thecarbonyl group. Tertiary butyl esters are particularly useful in this regard, because they are readilymade, t-butyl is extremely bulky, and yet they can can still be hydrolysed in aqueous acid under mildconditions by the method discussed on p. 000. In this example, deprotonation of t-butyl acetate withLICA (lithium isopropylcyclohexylamide) gives a lithium enolate that reacts with butyl iodide as thereaction mixture is warmed to room temperature.

Alkylation of carboxylic acidsThe lithium enolates of carboxylic acids can be formed if two equivalents of base are used. Carboxylicacids are very acidic so it is not necessary to use a strong base to remove the first proton but, since thesecond deprotonation requires a strong base such as LDA, it is often convenient to use two equivalentsof LDA to form the dianion. With carboxylic acids, even BuLi can be used on occasion because theintermediate lithium carboxylate is much less electrophilic than an aldehyde or a ketone.

The next alkylation of an acid enolate is of a carbamate-protected amino acid, glycine. As you sawin Chapter 25, carbamates are stable to basic reaction conditions. Three acidic protons are removedby LDA, but alkylation takes place only at carbon—the site of the last proton to be removed.Alkylation gets rid of one of the negative charges, so that, if the molecule gets a choice, it alkylates toget rid of the least stable anion, keeping the two more stabilized charges. A good alternative to usingthe dianion is to alkylate the ester or nitrile and then hydrolyse to the acid.


O

Ot-Bu

O

Ot-BuI

1. LICA, –78 °C

2.

85% yield

O

HO

OMe

OMe

IO

LiO

OMe

OMe

OLi

LiO

OMe

OMe

O

HO

OMe

OMe

HHBuLi BuLi

LWhy doesn’t BuLiadd to thecarboxylate as yousaw in Chapter 12to form theketone?Presumably in thiscase the aromaticring helps acidifythe benzylicprotons to tip thebalance towardsdeprotonation.Even withcarboxylic acids,LDA would be thefirst base youwould try.

P

You saw this sort of reactivity withdianions in Chapter 24: the lastanion to form will be the mostreactive.

BocHNOH

O

t-BuO NLi

OLi

O

O

t-BuO NLi

OLi

O

OPh

BrPh

Li

BocHNOH

O

Ph

LDA × 3

enolate trianion

H

•Alkylation of ketones, esters, and carboxylic acids is best carried out using thelithium enolates.

O

OR

O

OR

O

ORO

ORO O

OR

ORO

H H

O O

OR

base

ester enolatesecond molecule of

unenolized ester

self-condensation

the Claisen self-condensation of esters

acidic Hs

second molecule of base

Alkylation of aldehydesAldehydes are so electrophilic that, even with LDA at –78°C, the rate at which the deprotonationtakes place is not fast enough to outpace reactions between the forming lithium enolate and still-to-be-deprotonated aldehyde remaining in the mixture. Direct addition of the base to the carbonylgroup of electrophilic aldehydes can also pose a problem.

Using specific enol equivalents to alkylate aldehydes andketonesThese side-reactions mean that aldehyde enolates are not generally useful reactive intermediates.Instead, there are a number of aldehyde enol and enolate equivalents in which the aldehyde is pre-sent only in masked form during the enolization and alkylation step. The three most important ofthese specific enol equivalents are:

• enamines

• silyl enol ethers

• aza-enolates derived from imines

You met all of these briefly in Chapter 21, and we shall discuss how to use them to alkylate aldehy-des shortly. All three types of specific enol equivalent are useful not just with aldehydes, but withketones as well, and we shall introduce each class with examples for both types of carbonyl compound.

Enamines are alkylated by reactive electrophilesEnamines are formed when aldehydes or ketones react with secondary amines. The mechanism isgiven in Chapter 14. The mechanism below shows how they react with alkylating agents to form new

Using specific enol equivalents to alkylate aldehydes and ketones 671

Why do enolates alkylate on carbon?

Enolates have two nucleophilic sites: the carbon and the oxygen atoms: onp. 000 we showed that:

• Carbon has the greater coefficient in the HOMO, and is the softernucleophilic site

• Oxygen carries the greater total charge and is the harder nucleophilic site

In Chapter 21 you saw that hard electrophiles prefer to react at oxygen—that iswhy it is possible to make silyl enol ethers, for example. Some carbonelectrophiles with very good leaving groups also tend to react on carbon, but softelectrophiles such as alkyl halides react at carbon, and you will see only this typeof electrophile in this chapter.

R

O Me X

R

OMe

hard electrophiles react at O

X = OMs, OSO2OMe, +OMe2

R

O

Me X R

O

Me

soft electrophiles react at C

X = I, Br, Cl

In general:

• Hard electrophiles, particularly sulfates and sulfonates (mesylates,tosylates), tend to react at oxygen

• Soft nucleophiles, particularly halides (I > Br > Cl), react at carbon

• Polar aprotic solvents (HMPA, DMF) promote O-alkylation by separating the

enolate anions from each other and the counterion (making the bond morepolar and increasing the charge at O) while ethereal solvents (THF, DME)promote C-alkylation

• Larger alkali metals (Cs > K > Na > Li) give more separated ion pairs (morepolar bonds) which are harder and react more at oxygen

•Avoid using lithium enolates of aldehydes.

HH

O

R1

H

OLi

NiPr

iPr

R1

O

R1

Li

H

O

R1R1

LiOdeprotonation

lithium enolate

–78 °CTHF

aldol self-condensation

reactions which compete with aldehyde enolate formation

O

R1 Li

NiPr2

OLi

R1

NiPr2addition

–78 °CTHF

carbon–carbon bonds: the enamine here is the one derived from cyclohexanone and pyrrolidine.The product is at first not a carbonyl compound: it’s an iminium ion or an enamine (depending onwhether an appropriate proton can be lost). But a mild acidic hydrolysis converts the iminium ion orenamine into the corresponding alkylated carbonyl compound.

The overall process, from carbonyl compound to carbonyl compound, amounts to an enolatealkylation, but no strong base or enolates are involved so there is no danger of self-condensation.The example below shows two specific examples of cyclohexanone alkylation using enamines. Notethe relatively high temperatures and long reaction times: enamines are among the most reactive ofneutral nucleophiles, but they are still a lot less nucleophilic than enolates.

The choice of the secondary amine for formation of the enamine is not completely arbitrary eventhough it does not end up in the final alkylated product. Simple dialkyl amines can be used but cyclicamines such as pyrrolidine, piperidine, and morpholine are popular choices as the ring structure makesboth the starting amine and the enamine more nucleophilic (the alkyl groups are ‘tied back’ and can’tget in the way). The higher boiling points of these amines allow the enamine to be formed by heating.

α-Bromo carbonyl compounds are excellent electrophiles for SN2 reactions because of the rate-enhancing effect of the carbonyl group (Chapter 17). The protons between the halogen and the car-bonyl are significantly more acidic than those adjacent to just a carbonyl group and there is a seriousrisk of an enolate nucleophile acting as a base. Enamines are only very weakly basic, but react well asa nucleophile with a-bromo carbonyl compounds, and so are a good choice.

The original ketone here is unsymmetrical, so two enamines are possible. However, the formationof solely the less substituted enamine is typical. The outcome may be explained as the result of ther-modynamic control: enamine formation is reversible so the less hindered enamine predominates.


ONH

NR X

N

RH

N

R

N

R X

O

R

enamine

iminium enamine

1.

2. H2O, H+

mechanisms for the green steps are in Chapter 14

hydrolysis

cat H+

N

OBr1.

MeCN, reflux 13 h

2. H2O, 82 °C

N

OCl

Cl Cl1.

dioxane, reflux 22 h

2. HCl, H2O, 100 °C

NH

pyrrolidine

NH

piperidine

NH

O

morpholine

O

RBr

H H

NR2O

Ph

O

BrPh

O

O

R2NH

59% yield

1

2. H2Ocat. H+

For the more substituted enamine, steric hindrance forces the enamine to lose planarity, anddestabilizes it. The less substituted enamine, on the other hand, is rather more stable. Note howthe preference for the less substituted enamine is opposite to the preference for a more substitutedenol.

There is, however, a major problem with enamines: reaction at nitrogen. Less reactive alkylatingagents—simple alkyl halides such as methyl iodide, for example—react to a significant degree at Nrather than at C. The product is a quaternary ammonium salt, which hydrolyses back to the startingmaterial and leads to low yields.

That said, enamines are a good solution to the aldehyde enolate problem. Aldehydes form enam-ines very easily (one of the advantages of the electrophilic aldehyde) and these are immune to attackby nucleophiles—including most importantly the enamines themselves. Below are two examples ofaldehyde alkylation using the enamine method.


NRR

O NRRN

H

RR

H+

more hinderedenamine

less hinderedenamine

steric hindrance prevents complete planarity of conjugated enamine

N

R

N

R X

O

R

NN

Me I

OR

NMe

X X

R

OX

R

R X

H3O

H3O

iminium

quaternary ammonium salt starting material

+

alkylated carbonyl compound

simple alkyl halides do this

reactive alkylating agents do this

C -alkylation

N -alkylation

•Enamines can be used only with reactive alkylating agents.

• allylic halides

• benzyl halides

• αα-halo carbonyl compounds

Br

O

OEt

CO2Et

O

OHC

HN H

1. reflux MeCN

2. H2Ocat. H+ i -Bu2N

Both again use highly SN2-reactive electrophiles, and this is the main drawback of enamines. Inthe next section we consider a complementary class of enol equivalents that react only with highlySN1-reactive electrophiles.

Silyl enol ethers are alkylated by SN1-reactive electrophiles in the presence of LewisacidEnamines are among the most powerful neutral nucleophiles and react spontaneously with alkylhalides. Silyl enol ethers are less reactive and so require a more potent electrophile to initiate reac-tion. Carbocations will do, and they can be generated in situ by abstraction of a halide or other leav-ing group from a saturated carbon centre by a Lewis acid.

The best alkylating agents for silyl enol ethers are tertiary alkyl halides: they form stable carboca-tions in the presence of Lewis acids such as TiCl4 or SnCl4. Most fortunately, this is just the type ofcompounds that is unsuitable for reaction with lithium enolates or enamines, as elimination resultsrather than alkylation: a nice piece of complementary selectivity.

Below is an example: the alkylation of cyclopentanone with 2-chloro-2-methylbutane. Theketone was converted to the trimethylsilyl enol ether with triethylamine and trimethylsilylchloride:we discussed this step on p. 000 (Chapter 21). Titanium tetrachloride in dry dichloromethane pro-motes the alkylation step.

Aza-enolates react with SN2-reactive electrophilesEnamines are the nitrogen analogues of enols and provide one solution to the aldehyde enolateproblem when the electrophile is reactive. Imines are the corresponding nitrogen analoguesof aldehydes and ketones: a little lateral thinking should therefore lead you to expect some usefulreactivity from the nitrogen equivalents of enolates, known as aza-enolates. Aza-enolates are formedwhen imines are treated with LDA or other strong bases.

In basic or neutral solution, imines are less electrophilic than aldehydes: they reactwith organolithiums, but not with many weaker nucleophiles (they are more electrophilicin acid when they are protonated). So, as the aza-enolate forms, there is no danger at all of self-condensation.


N

OBr

CHONH

H1.

reflux MeCN

2. H2O

cat. H+

LYou saw the quantitative formation ofcarbocations by this method in Chapter17.

O

R Cl

SiMe3

O

SiMe3

R

Cl

O

RMe3SiCl

TiCl4 R

TiCl5

carbocation formed from alkyl halide

silyl enol ether

OSiMe3 Cl OO

TiCl4,CH2Cl2, 50 °C, 2.5 h

Me3SiCl, Et3NDMF, reflux

62% yield

The overall sequence involves formation of the imine from the aldehyde that is to be alkylated—usually with a bulky primary amine such as t-butyl- or cyclohexylamine to discourage even furthernucleophilic attack at the imine carbon. The imine is not usually isolated, but is deprotonated direct-ly with LDA or a Grignard reagent (these do not add to imines, but they will deprotonate them togive magnesium aza-enolates).

The resulting aza-enolate reacts like a ketone enolate with SN2-reactive alkylating agents—here,benzyl chloride—to form the new carbon–carbon bond and to re-form the imine. The alkylatedimine is usually hydrolysed by the mild acidic work-up to give the alkylated aldehyde.

In the next example, a lithium base (lithium diethylamide) is used to form the aza-enolate. Theease of imine cleavage in acid is demonstrated by the selective hydrolysis to the aldehyde without anyeffect on the acetal introduced by the alkylation step. The product is a mono-protected dialdehyde—difficult to prepare by other methods.


R1N

R2 R1N

R3

R1O

R1N

R1OH

R1O

aza-enolateimineenamine

enolatealdehydeenol

secondary amine, R2NHcat H+

primary amine, RNH2cat H+

base

not a stable species

stable, weakly nucleophilic

highly electrophilic

only weakly electrophilic

highly nucleophilic

highly nucleophilic

self-condensation a problem because of aldehydes’ electrophilicity

self-condensation not a problem because of imines’ low electrophilicity

base

P

Note. Aza-enolates are formedfrom imines, which can be madeonly from primary amines.Enamines are made fromaldehydes or ketones withsecondary amines.

N N

MgBr

HCHO

H2N

MgBr

cat H+

acidic proton

aza-enolate

aza-enolate formation

imine

strong base

Ph NN

MgBr

BnCl N

PhCl

Ph OMgBr

aza-enolate

THF, 23 h reflux

aza-enolate alkylation

H , H2O

CHOn -Bu

H2N

n -BuN

HR2N

n -BuN

Li

n -BuO

O

O

Br

O

O

Li

cat H+

LiNEt2,THF, –60 °C

imine1.

2. H+, H2O

Aza-enolate alkylation is so successful that it has been extended from aldehydes, where it isessential, to ketones where it can be a useful option. Cyclohexanones are among the most elec-trophilic simple ketones and can suffer from undesirable side-reactions. The imine from cyclohexa-none and cyclohexylamine can be deprotonated with LDA to give a lithium aza-enolate. In thisexample, iodomethylstannane was the alkylating agent, giving the tin-containing ketone afterhydrolysis.

Alkylation of ββ-dicarbonyl compoundsThe presence of two, or even three, electron-withdrawing groups on a single carbon atom makesthe remaining proton(s) appreciably acidic (pKa 10–15), which means that even mild bases canlead to complete enolate formation. With bases of the strength of alkoxides or weaker, only the mul-tiply stabilized anions form: protons adjacent to just one carbonyl group generally have apKa > 20. The most important enolates of this type are those of 1,3-dicarbonyl (or β-dicarbonyl)compounds.

The resulting anions are alkylated very efficiently. This diketone is enolized even by potassiumcarbonate, and reacts with methyl iodide in good yield. Carbonate is such a bad nucleophile that thebase and the electrophile can be added in a single step.


•Aldehyde alkylation

Aza-enolates are the best general solution for alkylating aldehydes with mostelectrophiles. With very SN2-reactive alkylating agents, enamines can be used, andwith very SN1-reactive alkylating agents, silyl enol ethers must be used

•Specific enol equivalents for aldehydes and ketones

To summarize:

• Lithium enolates can be used with SN2-reactive electrophiles, but cannot bemade from aldehydes

• Aza-enolates of aldehydes or ketones can be used with the same SN2-reactiveelectrophiles, but can be made from aldehydes

• Enamines of aldehydes or ketones can be used with allylic, benzylic, orαα-halocarbonyl compounds

• Silyl enol ethers of aldehydes or ketones can be used with SN1-reactive (ter-tiary, allylic or benzylic) alkyl halides

N N

Li

O

SnBu3

I SnBu3LDA,

THF

1.

2. H3O+

P

Typical electron-withdrawinggroups include COR, CO2R, CN,CONR2, SO2R, P=O(OR)2.

R

O O

H

R

O O

R

O O

Me

X

R

O O

MeHdelocalized enolate

alkylation of a 1,3-dicarbonyl compound (or β-dicarbonyl compound)

pKa 10–15pKa > 20:

not removedEtO

Among the β-dicarbonyls, two compounds stand out in importance—diethyl (or dimethyl) mal-onate and ethyl acetoacetate. You should make sure you remember their structures and trivialnames.

With these two esters, the choice of base is important: nucleophilic addition can occur at the estercarbonyl, which could lead to transesterification (with alkoxides), hydrolysis (with hydroxide), oramide formation (with amide anions). The best choice is usually an alkoxide identical with thealkoxide component of the ester (that is, ethoxide for diethy lmalonate; methoxide for dimethylmalonate). Alkoxides (pKa 16) are basic enough to deprotonate between two carbonyl groups but,should substitution occur at C=O, there is no overall reaction.

In this example the electrophile is the allylic cyclopentenyl chloride, and the base is ethoxide inethanol—most conveniently made by adding one equivalent of sodium metal to dry ethanol.

The same base is used in the alkylation of ethyl acetoacetate with butyl bromide.

Various electron-withdrawing groups can be used in almost any combination with good results.In this example an ester and a nitrile cooperate to stabilize an anion. Nitriles are not quite as anion-stabilizing as carbonyl groups so this enolate requires a stronger base (sodium hydride) in an aproticsolvent (DMF) for success. The primary alkyl tosylate serves as the electrophile.

Alkylation of â-dicarbonyl compounds 677

O O O O

Me I

O O

Me

K2CO3, MeI

acetone, reflux77% yield

LYou met these compounds, and theirstable enols, in Chapter 21.

EtO

O

OEt

O

EtO

OH

OEt

O

diethyl malonate stable enol tautomer

O

OEt

O OH

OEt

O

ethyl acetoacetate stable enol tautomer

HO

O

OH

O

malonic acid= propanedioic acid

O

OH

O

acetoacetic acid= 3-oxobutanoic acid

EtO2C CO2Et

EtO2C CO2Et

Cl

EtO

O O

OEtNaOEt, EtOH

61% yield

diethyl malonate

EtO2C

Br

EtO

O O

EtO

O O O

NaOEt, EtOH

BuBr

61% yield

ethyl acetoacetate

EtO2C

NC

TsOCO2Et

NC

NaH

DMF, pentane

+

These doubly stabilized anions are alkylated so well that it is common to carry out analkylation between two carbonyl groups, only to remove one of them at a later stage. This is madepossible by the fact that carboxylic acids with a b-carbonyl group decarboxylate (lose carbondioxide) on heating. The mechanism below shows how. After alkylation of the dicarbonylcompound the unwanted ester is first hydrolysed in base. Acidification and heating lead todecarboxylation via a six-membered cyclic transition state in which the acid proton is transferredto the carbonyl group as the key bond breaks, liberating a molecule of carbon dioxide. The initialproduct is the enol form of a carbonyl compound that rapidly tautomerizes to the more stableketo form—now with only one carbonyl group. Using this technique, β-keto-esters giveketones while malonate esters give simple carboxylic acids (both ester groups hydrolyse but onlyone can be lost by decarboxylation). Decarboxylation can occur only with a second carbonylgroup appropriately placed β to the acid, because the decarboxylated product must be formed as anenol.

The alkylation of ethyl acetoacetate with butyl bromide on p. 000 was done with the expressedintention of decarboxylating the product to give hexan-2-one. Here are the conditions for this decar-boxylation: the heating step drives off the CO2 by increasing the gearing on the entropy term (TDS‡)of the activation energy (two molecules are made from one).

Esters are much easier to work with than carboxylic acids, and a useful alternative procedureremoves one ester group without having to hydrolyse the other. The malonate ester is heated in a


MeO2C

O

R

O2C

O

R

O

R

O

R

O

OC

OH

R

O

O

H

HCl, heatNaOH, H2O

H

heat

decarboxylation of acetoacetate derivatives to give ketones

hydrolysis of ester

tautomerize

enol of ketone

ketone product

MeO2COMe

O

R

O2CO

O

R

OH

O

R

OH

O

R

O

OC

OH

OH

R

O

O

H

decarboxylation of malonate derivatives to give carboxylic acids

HCl, heatNaOH, H2O

H

heat

hydrolysis of ester

tautomerize

enol of carboxylic acid

carboxylic acid product

EtO2C

O

n -Bu

O2C

O

n -Bu

O

n -BuNa

NaOH, H2O HCl, heathexan-2-one61% yield

polar aprotic solvent—usually DMSO—in the presence of sodium chloride and a little water. Noacid or base is required and, apart from the high temperature, the conditions are fairly mild. Thescheme below shows a dimethyl malonate alkylation (note that NaOMe is used with the dimethylester) and removal of the methyl ester.

The mechanism is a rather unusual type of ester cleavage reaction. You met, in Chapter 17 andagain in Chapter 25, the cleavage of t-butyl esters in acid solution via an SN1 mechanism. In the reac-tion we are now considering, the same bond breaks (O–alkyl)—but not, of course, via an SN1 mech-anism because the alkyl group is Me. Instead the reaction is an SN2 substitution of carboxylate byCl–.

Chloride is a poor nucleophile, but it is more reactive in DMSO by which it cannot be solvated.And, as soon as the carboxylate is substituted, the high temperature encourages (entropy again) irre-versible decarboxylation, and the other by-product, MeCl, is also lost as a gas. The ‘decarboxylation’(in fact, removal of a CO2Me group, not CO2) is known as the Krapcho decarboxylation. Because ofthe SN2 step, it works best with methyl malonate esters.

We have only looked at single alkylations of dicarbonyl compounds, but there are twoacidic protons between the carbonyl groups and a second alkylation is usually possible. Excessof base and alkyl halide gives two alkylations in one step. More usefully, it is possible to intro-duce two different alkyl groups by using just one equivalent of base and alkyl halide in the firststep.

With a dihaloalkane, rings can be formed by two sequential alkylation reactions: this is an impor-tant way of making cycloalkanecarboxylic acids. Even the usually more difficult (see Chapter 42)four-membered rings can be made in this way.

Alkylation of â-dicarbonyl compounds 679

MeO2C

CO2Me

MeO2C

CO2Me

MeO2C

Br

+

NaOMe, MeOH NaCl

wet DMSO160 °C

92% yield 99% yield

O

O

Me ClMeO

O

R

O

O

MeO

O

R

MeCl CO2

MeO

O

R

HOH

MeO

O

RSN2

O

OH

acid-catalysed cleavageof t-butyl esters: SN1

OR

O

Nunormal nucleophilic

attack on C=O of ester

O

O

Me Cl

attack of Cl– on substituteddimethylmalonates: SN2

Br

MeO2C CO2Me MeO2C CO2Me

Br

MeO2C CO2MeNaH, THFNaH, THF

EtO2C CO2EtEtO2C CO2Et

Br Br

COOHCO2Et

Br

H

OEt

EtO

O

CO2Et

Br

EtO

O

+

excess NaOEt

1. NaOH2. HCl, heat

Ketone alkylation poses a problem in regioselectivityKetones are unique because they can have enolizable protons on both sides of the carbonyl group. Unlessthe ketone is symmetrical, or unless one side of the ketone happens to have no enolizable protons, tworegioisomers of the enolate are possible and alkylation can occur on either side to give regioisomeric prod-ucts. We need to be able to control which enolate is formed if ketone alkylations are to be useful.

Thermodynamically controlled enolate formationSelective enolate formation is straightforward if the protons on one side of the ketone are significant-ly more acidic than those on the other. This is what you have just seen with ethyl acetoacetate: it is aketone, but with weak bases (pKaH < 18) it only ever enolizes on the side where the protons are acid-ified by the second electron-withdrawing group. If two new substituents are introduced, in the man-ner you have just seen, they will always both be joined to the same carbon atom. This is an example ofthermodynamic control: only the more stable of the two possible enolates is formed.

This principle can be extended to ketones whose enolates have less dramatic differences in stabili-ty. We said in Chapter 21 that, since enols and enolates are alkenes, the more substituents they carrythe more stable they are. So, in principle, even additional alkyl groups can control enolate formationunder thermodynamic control. Formation of the more stable enolate requires a mechanism for equi-libration between the two enolates, and this must be proton transfer. If a proton source is available—and this can even be just excess ketone—an equilibrium mixture of the two enolates will form. Thecomposition of this equilibium mixture depends very much on the ketone but, with 2-phenylcyclo-hexanone, conjugation ensures that only one enolate forms. The base is potassium hydride: it’sstrong, but small, and can be used under conditions that permit enolate equilibration.


R

O

H

R

O

Me X

R

O

R

O

H

R

O

Me

R

O

MeMeX

H

H

B

B

regioisomeric enolates regioisomeric products

O

OEt

O

H

HH

H H

Me

O

OEt

O

Me

O

OEt

O

Me

O

OEt

O

H

HH

H Me

Me

O

OEt

O

Me

O

OEt

O

MeMe

R X

R

Me

O

Me

R

NaOEt

pKa ca. 12

pKa ca. 20

only more stable enolate forms

MeI

pKa ca. 12

pKa ca. 20NaOEt

1. NaOH2. HCl, heat

decarboxylatealkylate

introduction of bothnew substituents

directed by ester group

O

Ph

O

Ph

O

KH, THF

conjugated enolate formed

not formed

The more substituted lithium enolates can also be formed from the more substituted silyl enolethers by substitution at silicon—a reaction you met in Chapter 21. The value of this reaction nowbecomes clear, because the usual way of making silyl enol ethers (Me3SiCl, Et3N) typically produces,from unsymmetrical ketones, the more substituted of the two possible ethers.

One possible explanation for the thermodynamic regioselectivity in the enol ether-forming step isrelated to our rationalization of the regioselectivity of bromination of ketones in acid on p. 000.Triethylamine (pKaH 10) is too weak a base to deprotonate the starting carbonyl compound (pKa ca.20), and the first stage of the reaction is probably an oxygen–silicon interaction. Loss of a proton nowtakes place through a cationic transition state, and this is stabilized rather more if the proton being lostis next to the methyl group: methyl groups stabilize partial cations just as they stabilize cations.

An alternative view is that reaction takes place through the enol: the Si–O bond is so strong that evenneutral enols react with Me3SiCl, on oxygen, of course. The predominant enol is the more substitut-ed, leading to the more substituted silyl enol ether.

Kinetically controlled enolate formationLDA is too hindered to attack C=O, so it attacks C–H instead. And, if there is a choice of C–H bonds,it will attack the least hindered possible. It will also prefer to attack more acidic C–H bonds, and C–Hbonds on less substituted carbons are indeed more acidic. Furthermore, statistics helps, since a lesssubstituted C atom has more protons to be removed (three versus two in this example) so, even if therates were the same, the less substituted enolate would predominate.

Ketone alkylation poses a problem in regioselectivity 681

O

Me3SiCl

Et3N

OSiMe3 O

MeLi

Me3Si

Me Li

OLi

more substituted silyl enol ether

O Me3Si Cl OMe3Si

OMe3Si

H

OMe3Si

H

OSiMe3

NEt3

Et3N

Et3N

(+)

(+)

(+)(+)

‡ ‡

greater stabilization of cationic transition state by methyl group

cationic transition state not stabilized by methyl group

OOH OHMe3Si Cl NEt3

OMe3Si H

OSiMe3

major enol: more substitutedminor enol: less substituted

P

Think of base strengths: MeLi is aweaker base than t-BuLi, so theconjugate acid must be a strongeracid.

These factors multiply to ensure that the enolate that forms will be the one with the fewer sub-stituents—provided we now prevent equilibration of the enolate to the more stable, more substitutedone. This means keeping the temperature low, typically –78 °C, keeping the reaction time short, andusing an excess of strong base to deprotonate irreversibly and ensure that there is no remainingketone to act as a proton source. The enolate that we then get is the one that formed faster—thekinetic enolate—and not necessarily the one that is more stable.

In general, this effect is sufficient to allow selective kinetic deprotonation of methyl ketones, that is,where the distinction is between Me and alkyl. In this example, unusually, MeLi is used as a base:LDA was probably tried but perhaps gave poorer selectivity. The first choice for getting kinetic eno-late formation should always be LDA.

The same method works very well for 2-substituted cyclohexanones: the less substituted enolateforms. Even with 2-phenylcyclohexanone, which, as you have just seen, has a strong thermodynamicpreference for the conjugated enolate, only the less substituted enolate forms.

2-Methylcyclohexanone can be regioselectively alkylated using LDA and benzyl bromide by thismethod.


MeMe

O O

Me

O

H H H

H

H

NR2

add ketone toLDA, THF at –78 °C

less accessible(more hindered)

more accessible(less hindered)

kinetic enolate

P

There must never be more ketonein the mixture than base, orexchange of protons betweenketone and enolate will lead toequilibration. Kinetic enolateformations with LDA must bedone by adding the ketone to theLDA so that there is excess LDApresent throughout the reaction.

O O

PhBr Ph

1. MeLi, HMPA, Et2O

2.77% yield

OLi

Ph

O

Ph

OLi

PhHH H

100% formed

more accessible (less hindered)

add to LDA, –78 °C

0% formed

OLi

Me

O

MeHH H

O

PhBrPh

99% formed

add to LDA, –78 °C

45% yield

•Regioselective formation of enolates from ketones

R1

O

R1

O

R1

O

H H kinetic enolatethermodynamic enolate

base base

Thermodynamic enolates are:

• more substituted

• more stable

• favoured by excess ketone, hightemperature, long reaction time

Kinetic enolates are:

• less substituted

• less stable

• favoured by strong, hindered base,low temperature, short reaction time

Enones provide a solution to regioselectivity problemsEnolates can be made regiospecifically from, for example, silyl enol ethers or enol acetates just bytreating them with an alkyllithium. These are both substitution reactions in which RLi displaces theenolate: one is SN2(Si) and the other is attack at C=O. Provided there is no proton source, the eno-late products have the same regiochemistry as their stable precursors, and single enolate regioiso-mers are formed. But there is a problem: forming enol ethers or enol esters will usually itself requirea regioselective enolization! There are two situations in which this method is nonetheless useful:when the more substituted lithium enolate (which is hard to make selectively otherwise) is required,and when a silyl enol ether can be formed by a method not involving deprotonation. These methodsare what we shall now consider.

Dissolving metal reduction of enones gives enolates regiospecificallyIn Chapter 24 you met the Birch reduction: the use of dissolving metals (K,Na, or Li in liquid ammonia, for example) to reduce aromatic rings andalkynes. The dissolving metal reduction of enones by lithium metal in liquidammonia is similar to these reactions—the C=C bond of the enone isreduced, with the C=O bond remaining untouched. An alcohol is required asa proton source and, in total, two electrons and two protons are added in astepwise manner giving net addition of a molecule of hydrogen to the doublebond.

Enones provide a solution to regioselectivity problems 683

Dianions allow unusual regioselectivity in alkylations of methyl acetoacetate

In Chapter 24, we introduced the idea that the last-formed anion in a dianion ortrianion is the most reactive. Methyl acetoacetate is usually alkylated on thecentral carbon atom because that is the site of the most stable enolate. Butmethyl acetoacetate dianion—formed by removing a second proton from the

usual enolate with a very strong base (usually butyllithium)—reacts first on theless stable anion: the terminal methyl group. Protonation of the more stableenolate then leads to the product. Butyllithium can be used as a base becausethe anionic enolate intermediate is not electrophilic.

OMe

OO

OMe

OO

OMe

OO

H

Bu LiOMe

OO

OMe

OOBr

OMe

OLi

OMe

OO

Li

Li

NaH BuLi

dianion

79% yield

HO

R

OSiMe3

R

OLi

R

OSiMe3

Me Li

MeLi + SiMe4

silyl enol ether lithium enolate

R

OAc

R

OLi

R

O

Me LiO

O Me LiMeLi × 2

+

enol acetate lithium enolate

OLi

O OH

H

Li, NH3

ROH

C=C reduced

The mechanism follows that described on p. 000: transfer of an electron forms a radical anion thatis protonated by the alcohol to form a radical. A second electron transfer forms an anion that canundergo tautomerization to an enolate.

The enolate is stable to further reduction, and protonation during the work-up will give a ketone.But reaction with an alkyl halide is more fruitful: because the enolate forms only where the doublebond of the enone was, regioselective alkylation becomes possible.

You saw above that an equilibrium mixture of the enolates of 2-methylcyclohexanone containsonly about a 4:1 ratio of regioisomers. By reducing an enone to an enolate, only 2% of the unwantedregioisomer is formed.

The transfer of electrons is not susceptible to steric hindrance so substituted alkenes pose noproblem. In the next example, the enolate reacts with allyl bromide to give a single stereoisomer ofthe product (the allyl bromide attacks from the face opposite the methyl group). Naturally, only oneregioisomer is formed as well, and it would be a tall order to expect formation of this single enolateregioisomer by any form of deprotonation method.

Conjugate addition to enones gives enolates regiospecificallyAlthough we did not talk in detail about them at that time, you will recall from Chapter 10 that con-jugate addition to enones generates first an enolate, which is usually protonated in the work-up. But,again, more fruitful things can be done with the enolate under the right conditions.


O O OH OH OH

e EtOH e

enolate

OO O

R

IRno alkylation here

alkylation here

H2O, H+

O OLi O O

+Li, NH3, ROH MeI

60% 2%one regioisomer

O OLi O

BrLi, NH3, ROH

45% yield

(±)

R

O

R

O

NuNu

R

O

Nu

R

O

Nu

E

H

enolate E

conjugate addition

direct addition must be avoided

The simplest products are formed when Nu = H, but this poses a prob-lem of regioselectivity in the nucleophilic attack step: a nucleophilichydride equivalent that selectively undergoes conjugate addition to theenone is required. This is usually achieved with extremely bulky hydridereagents such as lithium or potassium tri(sec-butyl)borohydride (oftenknown by the trade names of L- or K-Selectride, respectively). In this exam-ple, K-Selectride reduces the enone to an enolate that is alkylated by methyliodide to give a single regioisomer. The reaction also illustrates the differ-ence in reactivity between conjugated and isolated double bonds.

With organocopper reagents, conjugate addition introduces a new alkyl group and, if the result-ing enolates are themselves alkylated, two new C–C bonds can be formed in a single step (a tandemreaction: one C–C bond-formation rides behind another). In Chapter 10 we explained that the bestorganocuprate additions are those carried out in the presence of Me3SiCl: the product of these reac-tions is a silyl enol ether, formed regioselectively (the ‘enol’ double bond is always on the side wherethe enone used to be).

The silyl enol ethers are too unreactive for direct alkylation by an alkyl halide, but by convertingthem to lithium enolates all the usual alkylation chemistry becomes possible. This type of reactionforms the key step in a synthesis of the natural product α-chamigrene. Conjugate addition ofMe2CuLi gives an enolate that is trapped with trimethylsilyl chloride. Methyllithium converts theresulting silyl enol ether into a lithium enolate (by SN2 at Si). The natural product has a spiro six-membered ring attached at the site of the enolate, and this was made by alkylating with a dibromide(you saw this done on p. 000). The first substitution is at the more reactive allylic bromide. A secondenolization is needed to make the ring, but this can be done under equilibrating conditions becausethe required six-membered ring forms much faster than the unwanted eight-membered ring thatwould arise by attack on the other side of the ketone.

Enones provide a solution to regioselectivity problems 685

H BM M = Li: lithium tri-sec-butyl-borohydride (L-Selectride)

M = K: potassium tri-sec-butyl-borohydride (K-Selectride)

bulky reducing agents

O OO

BHs-Bu

s-Bu s-Bu

OK-SelectrideTHF

–78 °CMeI, THF

–78 °C to 0 °C98% yield

O O

MeMe

OSiMe3

Me

O

Me

RCuMe

OMe3SiCl

RX

conjugate additionMe2CuLi

hard electrophile attacks O

soft electrophile attacks C

O

Me

OSiMe3 OBr

Br

Me

Me

OLi

O

Br

KH,

THF, 25 °C

MeLi

THF

Me2CuLi, Me3SiCl

78% yield

Among the most important of these tandem conjugate addition–alkylation reactions are those ofcyclopentenones. With cyclopentenone itself, the trans diastereoisomer usually results because thealkylating agent approaches from the less hindered face of the enolate.

This is the sort of selectivity evident in the next example, which looks more complicated but is reallyjust addition of an arylcopper reagent followed by alkylation (trans to the bulky Ar group) with aniodoester.

One of the most dramatic illustrations of the power of conjugate addition followed byalkylation is the short synthesis of the important biological molecule prostaglandin E2 by RyojiNoyori in Japan. The organocopper reagent and the alkylating agent contain all the function-ality required for both side chains of the target in protected form. The required trans stereo-chemistry is assembled in the key step, which gives a 78% yield of a product requiring onlyremoval of the silyl ether and ester protecting groups. The organometallic nucleophile wasprepared from a vinyl iodide by halogen–metal exchange (Chapter 9). In the presence of copperiodide this vinyllithium adds to the cyclopentenone in a conjugate sense to give an intermediate eno-late. Because in this case the starting enone already has a stereogenic centre, this step is also stereo-selective: attack on the less hindered face (opposite the silyl ether) gives the trans product. Theresulting enolate was alkylated with the allylic iodide containing the terminal ester: once again thetrans product was formed. It is particularly vital that enolate equilibration is avoided in this reactionto prevent the inevitable E1cB elimination of the silyloxy group that would occur from the otherenolate.


O OLi

Me

O

Me

R

O

Me

RMe2CuLi RX

alkylation from less hindered face

minormajor

(±)(±)

MeO

CuR

MeO

O

EtO2C

IEtO2C

O

Ar

O

MgBr

95% yield

LRyoji Noyori works at the University ofNagoya in Japan. He has introducedmany methods for making molecules,the most important of which allow theformation of single enantiomers usingchiral catalysts. You will meet somemore of his chemistry in Chapter 45.

OSiR3R3SiO

O

CO2MeI CO2Me

C5H11

OLi

R3SiOOSiR3

OSiR3 R3SiO O

C5H11I Li

OSiR3

OHOH

O

CO2H

BuLi

1. Bu4NF

2. H+

78% yield

CuI1.

2.

trans

trans

prostaglandin E2

To conclude...We have considered the reactions of enolates and their equivalents with alkyl halides. In the nextchapter we move on to consider the reactions of the same types of enolate equivalents with a differ-ent class of electrophiles: carbonyl compounds themselves.

Summary of methods for alkylating enolates

Specific enol equivalent Notes

To alkylate esters

• LDA → lithium enolate

• use diethyl- or dimethylmalonate and decarboxylate gives acid (NaOH, HCl) or ester (NaCl, DMSO)

To alkylate aldehydes

• use enamine with reactive alkylating agents

• use silyl enol ether with SN1-reactive alkylating agents

• use aza-enolate with SN2-reactive alkylating agents

To alkylate symmetrical ketones

• LDA → lithium enolate

• use acetoacetate and decarboxylate equivalent to alkylating acetone

• use enamine with reactive alkylating agents

• use silyl enol ether with SN1-reactive alkylating agents

• use aza-enolate with SN2-reactive alkylating agents

To alkylate unsymmetrical ketones on more substituted side

• Me3SiCl, Et3N → silyl enol ether with SN1-reactive alkylating agents

• Me3SiCl, Et3N → silyl enol ether → lithium enolate with SN2-reactive alkylating agentswith MeLi

• alkylate acetoacetate twice and decarboxylate two successive alkylations of ethyl acetoacetate

• addition or reduction of enone to give specific lithiumenolate or silyl enol ether

To alkylate unsymmetrical ketones on less substituted side

• LDA → kinetic lithium enolate with SN2-reactive electrophiles

• LDA then Me3SiCl → silyl enol ether with SN1-reactive electrophiles

• use dianion of alkylated acetoacetate and decarboxylate two successive alkylations of ethyl acetoacetate

• use enamine with reactive electrophiles

To conclude . . . 687

1. Suggest how the following compounds might be made by thealkylation of an enol or enolate.

2. And how might these compounds be made using alkylation ofan enol or enolate as one step in the synthesis?

3. And, further, how might these amines by synthesized usingalkylation reactions of the enolate style as part of the synthesis?

4. This attempted enolate alkylation does not give the requiredproduct. What goes wrong? What products would be expectedfrom the reaction?

5. Draw mechanisms for the formation of this enamine, itsreaction with the alkyl halide shown, and the hydrolysis of theproduct.

6. How would you produce specific enol equivalents at the pointsmarked with the arrows (not necessarily starting from the simplecarbonyl compound shown)?

7. How would the reagents you have suggested in Problem 6 reactwith: (a) Br2; (b) a primary alkyl halide RCH2Br?

8. Draw a mechanism for the formation of the imine fromcyclohexylamine and the following aldehyde.

9. How would the imine from Problem 8 react with LDA followedby n-BuBr? Draw mechanisms for each step: reaction with LDA,reaction of the product with n-BuBr, and the work-up.

10. What would happen if thisshort cut for the reaction inProblems 8 and 9 were tried?

11. Suggest mechanisms for these reactions.

12. How does this method of making cyclopropyl ketones work?Give mechanisms for all the reactions.

13. Give the structures of the intermediates in the followingreaction sequence and mechanisms for the reactions. Commenton the formation of this particular product.

14. Suggest how the following products might be made usingenol or enolate alkylation as at least one step. Explain your choiceof specific enol equivalents.


Problems

EtO2C CO2Et

O

O OO

O

NH2

NH2

R

R

Me CHO CHO1. BuLi

2. i-PrCl×

ONH

O

N

O

Br

O

N

O

O

O

O

H

H

cat.

H2O

OO

O

RCHO

H2N R NHcat.

R N

1. LDA

2. BuBr?

RCHO

1. LDA

2. BuBr?

Br BrEtO2C

EtO2C

CO2H1.

2. NaOH, H2O 3.

EtO , EtOH

H heat

CO2Et

O

O O

O O

Br

O O

EtO , EtOH

HBr base

NH

O

N

OMe

1. NaNH22. MeOTs

3. LDA4. EtBr

Ph

CHO

OCO2H

building on: arriving at - chtf.stuba.skszolcsanyi/education/files/organicka chemia ii... · meo me...

Documents