Jones M., Fleming S.A. Organic Chemistry

Подождите немного. Документ загружается.

21.4 Single Bonds as Neighboring Groups 1109

the classic mechanistic question: Is there a second intermediate—in this case the

secondary carbocation—or is the reaction a concerted formation of the tertiary car-

bocation (Fig. 21.46)?

PROBLEM 21.21 Draw Energy versus Reaction progress diagrams for the two pos-

sibilities of Figure 21.46. How many energy barriers separate starting material

from the tertiary cation in each mechanism?

The question of the number of steps in the mechanism has been answered in this

system by the observation of an isotope effect.The rates of reaction of the unlabeled

tosylate and the deuterated tosylate of Figure 21.47 are measured. If the mechanism

involves slow formation of the secondary carbocation, followed by a faster hydride

shift to give the more stable tertiary carbocation,substitution of deuterium for hydro-

gen would have no great effect on the rate. Even though the carbon–deuterium bond

is stronger than the carbon–hydrogen bond,it is not broken in the rate-determining step

of the reaction. By contrast, if the hydride (or deuteride) shifts as the leaving group

departs, the carbon–hydrogen or carbon–deuterium bond is broken in the rate-deter-

mining step and a substantial effect on the rate would be expected.

OTs

H(D)

2. deprotonate

–

OTs

Found

> 2.2

Predicted

⬇ 1.1

H(D)

OTs

H(D)

fastslow

(a)

(b)

slow

H(D)

1. CH

COOH

2. deprotonate

1. CH

COOH

FIGURE 21.47 (a) A small isotope

effect is predicted for the two-step

mechanism. (b) There can only be a

substantial isotope effect if breaking

the carbon–hydrogen or

carbon–deuterium bond occurs in

the rate-determining ionization step.

A larger isotope effect means that

the carbon–hydrogen or

carbon–deuterium bond must be

breaking in the ionization step of

the reaction.

The rate for the unlabeled compound was found to be more than twice as fast

as that for the deuterated tosylate.The measured isotope effect (k

= 2.2) reveals

that the hydride must move as the leaving group departs. If the slow step in the reac-

tion had involved a simple ionization to a secondary carbocation, followed by a

hydride shift, only a very small effect of the label would have been observed (prece-

dent suggests k

= 1.1 to 1.3). The observation of a substantial isotope effect

shows that the reaction mechanism must include a neighboring group effect, this

time of a neighboring sigma bond!

Now comes an even more subtle question.Given that even single bonds can par-

ticipate in ionization reactions, are cyclic, delocalized intermediates possible?

PROBLEM 21.22 Although it is hard to see why now, phenonium ions (p. 1098)

were originally very controversial. Despite their evident similarity to the interme-

diates of aromatic substitution (Fig. 21.27), there was fierce opposition to the

notion of their existence. Explain why a confusion between neighboring π partic-

ipation and neighboring σ participation may have led to a misunderstanding.

1110 CHAPTER 21 Intramolecular Reactions and Neighboring Group Participation

A cyclic ion

PROBLEM 21.23 There has been something of a controversy over the structure of

the ethyl cation, CH

. Contrast orbital pictures of the straightforward open

ion with those of a delocalized version, which you might consider making

through protonation at the center of the double bond of ethylene. Make analo-

gies! Where have you seen a triangular array of molecular orbitals recently?

xo-2-Norbornyl tosylate

Relative rate ~ 350

exo-2-Norbornyl acetate endo-2-Norbornyl tosylat

Relative rate = 1

OTs

COOH CH

COOH

OTs

WEB 3D WEB 3D

FIGURE 21.48 Both endo- and exo-2-

norbornyl tosylate react in acetic acid

to give exo-2-norbornyl acetate.The

reaction of the exo isomer is faster.

The exo compound reacted at a faster rate than the endo compound, with the

exact difference depending on reaction conditions. At first sight this rate increase

would seem to provide evidence for neighboring group participation, but the oppo-

nents of participation of σ bonds quite rightly pointed out that the rate accel-

eration was small ( 350) and suggested that this effect could be the result of other

differences between the exo and endo tosylates. This argument brings up a most

important point. Whenever you are given a rate difference, you must ask yourself,

“To what reaction does this rate difference refer?” In this case, the factor of 350

referred to the formation of acetate product. Is that what we are really interested in?

Certainly not.When we speak of neighboring group participation, we mean partic-

ipation in ion formation, and that step is not necessarily equivalent to the product

formation step. If ion formation is faster then we need a way to measure the rate

acceleration in the ionization step, not the product-forming step (Fig. 21.49).

fast

ionization step

OTs

1. CH

COOH

2. deprotonation

slower

product-forming step

FIGURE 21.49 If the rate difference is determined by observing formation of acetate, we may be missing a fast, reversible

ionization step.

The most famous argument of modern times in organic chemistry concerned

the question of delocalized intermediates produced by σ bond participation,

and a fierce one it was—and still sometimes is. The reaction in question was the

solvolysis (S

1) reaction of the 2-norbornyl tosylates,the tosylates of exo- and endo-

bicyclo[2.2.1]heptan-2-ol (Fig. 21.48).

Fast step

Ionization

Slow step

Product-forming

steps

Energy

Reaction progress

Tosylate

OTs

Racemic

acetate

OAc

Ion

21.4 Single Bonds as Neighboring Groups 1111

Perhaps an Energy versus Reaction progress diagram makes the point more

clearly. If we measure the rate of appearance of acetate (green), we miss the faster

ionization reaction (red) (Fig. 21.50).

FIGURE 21.50 An Energy versus

Reaction progress diagram for the

1 solvolysis of exo-2-norbornyl

tosylate. A rapid ionization precedes

a slower formation of acetate.

OTs

deprotonate

OAc

exo Acetate—retention!

and racemization!

exo Tosylate (

= optically active)

Optically active endo acetate

(not formed)

OAc

AcO

COOH

FIGURE 21.51 A simple mechanism involving S

2 displacement by acetic acid predicts optically active endo acetate, which

is not what happens.The product is exo acetate. And it is racemic!

The stereochemistry of the reaction also points to neighboring group participa-

tion. The product from the optically active exo tosylate is not that of a simple S

reaction, which would be the inverted optically active endo acetate (Fig. 21.51).

Instead, it is the product of retention, the exo acetate. We have been alerted to look

for just this kind of stereochemical result as evidence of a neighboring group effect—

it’s clue 1. Moreover, optically active tosylate gives racemic acetate. We will see how

this is evidence of an apparent rearrangement, which is clue 2.

1112 CHAPTER 21 Intramolecular Reactions and Neighboring Group Participation

PROBLEM 21.24 Draw resonance forms for the delocalized ion of Figure 21.52.

It was also quite rightly pointed out that the delocalized ion not only explained

the observed retention of stereochemistry (exo tosylate gave exo acetate), but

demanded a racemic product, because it is achiral (Fig. 21.53). The bridged ion

doesn’t look symmetrical at first, but if you turn it over, you can see that it clearly is.

OTs

OAc

exo Tosylate exo Acetate

Delocalized ion

Intermolecular

Intramolecular

Deprotonation

FIGURE 21.52 If the carbon–carbon

σ bond assists in ionization

(a neighboring group effect), a

cyclic, delocalized ion can result.

Opening of this ion by acetic acid,

followed by deprotonation, would

give the observed stereoisomer of

the product.

OTs

Optically active Achiral!

Use this dotted

carbon to orient

yourself

FIGURE 21.53 The observed

formation of racemic product (as well

as the racemization of recovered

starting tosylate) can be explained by

the achiral nature of the cyclic ion.

Addition of acetic acid must occur with equal probability at the two equivalent car-

bons of the delocalized ion to give both enantiomers of exo-2-norbornyl acetate (Fig.21.54).

Proponents of neighboring group participation asserted that these pieces of stereo-

chemical data constituted powerful evidence for a delocalized ion. First of all, it is

only the exo tosylate that can form such an ion directly, and the exo compound did

indeed react at a greater rate than the endo isomer (Fig. 21.52).

AcO

OAc

Enantiomers!

OAc

deprotonate deprotonateadd

add

FIGURE 21.54 Addition of

acetic acid to the delocalized

ion must give a pair of

enantiomers in equal

amounts—a racemic mixture.

This notion also suggests how to get a better measure of the rate of ionization

than simply measuring the rate of acetate formation. If the initial ionization is

reversible, then starting tosylate will be racemized as soon as the leaving group has

21.4 Single Bonds as Neighboring Groups 1113

TsO

OTs

An achiral ion

Enantiomers

Racemic

tosylate

H H

OTs

–

OTs

reversible

ionization

–

OTs

WEB 3D

FIGURE 21.55 If the leaving group

(tosylate ion) re-adds to the cyclic ion

to re-form starting material,

racemization results.

So, we should be able to get a better measure of the rate of ionization by deter-

mining the rate of racemization of starting tosylate.This clever experiment has now

been done and reveals a rate increase of about 1500 for the exo tosylate over the endo,

a more impressive rate acceleration than the originally reported factor of 350.

Now we have three traditional kinds of evidence for neighboring group partic-

ipation: rate acceleration,apparent rearrangement,and retention of stereochemistry.

It looks like a strong case can be made for neighboring group participation. What

could the opponents of the three-carbon, two-electron system say?

First of all, they pointed out that the rate acceleration was small (they preferred

to focus on the original factor of 350, but even 1500 isn’t very large), and they were

right. Moreover, one could invoke assistance by the σ bond in a simple way, with-

out forming a cyclic ion. That would give some rate acceleration in the exo mole-

cule, as there can be no assistance in the endo isomer (Fig. 21.56).

fast

equilibrium

exo

endo

no frontside

OTs

Enantiomers

FIGURE 21.56 Formation of two,

classical, rapidly equilibrating ions

can explain both the fast rate of the

exo isomer and the racemization.

The classicists actually argued most strongly that the rate of ionization of the endo compound was especially

slow and that the exo compound was more or less normal.

sufficiently departed the cation (Fig. 21.55).The tosylate needs to be far enough from

the norbornyl framework so that it can’t tell which carbon it came from. It needs to

become symmetrically disposed with respect to that carbon.

How would such a species explain the stereochemical results: racemization and

retention of the exo configuration? Racemization was easy, one had only to permit

the initially formed ion to equilibrate rapidly with the equivalent,mirror-image form

as noted in Figure 21.56.

Always keep in mind the difference between the equilibrating open ions of Figure

21.56 and the delocalized ion. The delocalized ion is a single species. But be care-

ful; the resonance forms of the bridged ion do look like the separate, equilibrating

forms postulated by the classicists (Fig. 21.57).

1114 CHAPTER 21 Intramolecular Reactions and Neighboring Group Participation

OTs

Delocalized ion Resonance forms

Two equilibrating structures

FIGURE 21.57 The two mechanistic proposals compared.

OAc

OTs

COOH

OTs

TsO

rotate

90⬚

WORKED PROBLEM 21.25 Explain the formation of exo-2-norbornyl acetate from

the following reaction:

ANSWER This reaction is just another way of forming the 2-norbornyl cation.

Symmetrical displacement of the leaving group by the double bond of the five-

membered ring gives the ion.

(continued)

21.4 Single Bonds as Neighboring Groups 1115

2 deprotonate

OAc

Opening by acetic acid in an intermolecular S

2 reaction must produce the exo

acetate.

How did the classicists explain retention: the exclusive formation of exo acetate?

That task was much harder, and to our taste, never quite achieved.The two faces of

the molecule are different,however, and one must admit that in principle, there might

be sufficient difference between them to permit the exclusive formation of the exo

acetate (Fig. 21.58). Much chemistry of the norbornyl system reveals a real prefer-

ence for exo reactions, and the question again became one of degree.How much exo

addition could be demanded by an open, localized ion?

exo

–

Attack at the endo face

gives endo product

Attack at the exo face

gives exo product

endo

–

FIGURE 21.58 The exo and endo

faces of the open ion are different.

Addition of a nucleophile could,

in principle, occur from only the

exo side.

In recent years, an increasing number of sophisticated spectroscopic techniques

have been applied to this long-standing controversy. The 2-norbornyl cation has

actually been detected by NMR spectroscopy at very low temperature in a highly

polar but nonnucleophilic solvent. Another spectroscopic technique, pioneered by

Martin Saunders of Yale University (b. 1931), focuses on the general differences in

the

C NMR spectra between equilibrating structures and delocalized molecules.

For example, the deuteriolabeled cyclohexenyl cation in Figure 21.59, certainly a

resonance-stabilized species, shows essentially no difference between the

C NMR

signals for the two carbons that share the positive charge.

superacid

XD HD

In the

C NMR spectrum,

the red-dot carbons are

very nearly equivalent

FIGURE 21.59 In this resonance-

stabilized carbocation, the two “red-

dot” carbons sharing the positive

charge are not strongly differentiated

in the

C NMR spectrum by

deuterium substitution.

1116 CHAPTER 21 Intramolecular Reactions and Neighboring Group Participation

fast

equilibration

The delocalized structure is favored

OTs

Very similar signals

in the

C NMR spectrum

The two red-dot carbons should be widely separated in the

C NMR spectrum

FIGURE 21.60 The

C NMR

spectrum shows only a tiny

difference between the

deuterated and undeuterated

positively charged carbons.

The 2-norbornyl cation

is delocalized.

Saunders demonstrated that in equilibrating structures very great differences, on

the order of 100 ppm, appear between such carbons. The Saunders technique gets

right to the heart of the matter—to the center of the dispute over the 2-norbornyl

cation: Is it a single, resonance-stabilized delocalized structure, or an equilibrating

set of independent ions? Use of Saunders’ technique in this case is complicated by

the presence of hydride shifts in the 2-norbornyl cation, but nonetheless the weight

of the evidence seems to favor strongly the delocalized structure (Fig. 21.60).

An equilibrating pair of tertiary cations

superacid

FIGURE 21.61 Not all ions are

delocalized. This related but tertiary

carbocation is not bridged.

Are all bicyclic cations delocalized? For a while it certainly seemed so. At least they

were drawn that way during what might be called the rococo phase of this theory. But

there is no reason why they should all be delocalized.It is one form of stabilization,noth-

ing more,and if other forms of stabilization are available,ions will be open,classical species.

For example,the classicists were right when they argued that certain tertiary carbocations

are localized. Tertiary carbocations are more stable than secondary ones, and so there is

no need for the special stabilization of three-center, two-electron bonding. The propo-

nents of delocalization should not have been upset, but they certainly were (Fig. 21.61).

The argument over the nature of the 2-norbornyl cation and related molecules was/is remarkable not only for its

longevity, but for its intensity as well. The argument became very nasty and both sides definitely lost their cool.

Despite the common notion of scientists as Dispassionate Seekers after Truth, we are all too prone to the human

vices, in this case overdefending our ideas (our intellectual children). In one corner were the proponents of three-

center,two-electron bonding,chief among them Saul Winstein and Donald Cram.In the other was H.C. Brown,

who passionately maintained that all could be explained through conventional, localized intermediates.This argu-

ment was no battle of lightweights! Brown and Cram have since won the Nobel prize (for other work),and Winstein

should have been so honored—and would have been,had he lived long enough.The dispute was raucous and fea-

tured many a shouting match at meetings and seminars. It is not easy to be insulting in the body of a scientific

paper, because the referees (two or more readers who judge whether the work is publishable) will not stand for it.

However, it is much easier to be nasty in the footnotes! There are some really juicy footnotes in the literature on

this subject. In our view, both sides went overboard. Although there is much virtue in avoiding new, unwarranted

explanations, the classicists were much too resistant to change.There is nothing inherently crazy about a delocal-

ized ion! Three-center, two-electron bonding is a fine way to hold molecules together and is quite common in

inorganic chemistry, and even parts of organic chemistry. On the other hand, the delocalizers overdefended their

ground. Just because some ions are delocalized in this fashion doesn’t mean that all are!

21.5 Coates’ Cation 1117

OSO

OAc

COOH

(AcOH)

WEB 3D

FIGURE 21.62 The solvolysis of this

sulfonate in acetic acid gives the

related acetate.

21.5 Coates’ Cation

Robert M. Coates (b. 1938) of the University of Illinois has found an ion for which

there is unequivocal evidence for a delocalized structure. The ion cannot be repre-

sented by a rapidly equilibrating set of localized structures and reacts in acetic acid

to give an acetate as shown in Figure 21.62.

OSO

View from

here...

...to get

this diagram

FIGURE 21.63 Participation of the

carbon–carbon σ bond to give a

symmetrical delocalized ion. Notice

that three different kinds of carbons

exist.The

C NMR spectrum will

consist of three lines.

At low temperature in superacid, ionization with neighboring group participa-

tion of the cyclopropane σ bond gives a delocalized cation that must show three sig-

nals in its

C NMR spectrum (Fig. 21.63).

In this molecule,an equilibrating set of ions must give quite a different

C NMR

spectrum.

etc.

Equilibrating open ions

OSO

PROBLEM 21.26 How many signals would a rapidly equilibrating “Coates’ cation”

show in its

C NMR spectrum?

Happily, the sulfonate ionizes to give a stable ion at low temperature in superacid

and the case is settled, because the spectrum consists of exactly the three signals

required for a delocalized structure: no more and no fewer. Coates’cation is a three-

carbon, two-electron system, there is no doubt about it.

1118 CHAPTER 21 Intramolecular Reactions and Neighboring Group Participation

Key Terms

anchimeric assistance (p. 1089)

episulfonium ion (p. 1088)

isotope effect (p. 1109)

neighboring group effect (p. 1081)

norbornyl system (p. 1099)

phenonium ion (p. 1098)

three-center, two-electron bonding

(p. 1102)

Reactions, Mechanisms, and Tools

This chapter features intramolecular displacement using a

variety of internal nucleophiles. Some, such as the heteroatoms

bearing lone pairs of electrons, are easy to understand, and really

just recapitulate intermolecular chemistry (Fig. 21.2).

Other nucleophiles, such as the electrons in π or σ bonds,

are less obvious extensions of old chemistry. Here, the nucleo-

philes involved are very weak and there often is not a clear anal-

ogy to chemistry you already know. One case in which there is a

good connection is formation of phenonium ions through the

action of a benzene ring as a neighboring group. The analogy is

to the intermolecular reaction of benzenes with electrophiles in

the aromatic substitution reaction (Figs. 21.26 and 21.27).

Displacement by carbon–carbon π or σ bonds can lead

to delocalized ions about which there has been much

controversy. It now appears that such intermediates are

involved in some ionizations, and that the three-center, two-

electron bonding required is an energetically favorable

situation, at least for highly electron-deficient systems such

as carbocations (Fig. 21.57).

Common Errors

The problem with this material is mostly one of recognition.

Once it is clear that you have a neighboring group problem,

essentially every difficulty is resolved by a pair of S

2 reactions:

the first intramolecular, the second intermolecular. Search for the

internal nucleophile, and use it to displace the leaving group.

There can be no denying that many of the structures

encountered in neighboring group problems are complex and

sometimes hard to untangle. A useful technique is to draw the

result of the arrow formalism without moving any atoms, then

to relax the structure, probably containing long or even bent

bonds, to a more realistic picture. Trying to do both steps at

once is dangerous.

A minor problem appears in recognizing that in some delo-

calized ions, the web of dashes representing three-center, two-

electron bonding can operate as a leaving group. It is easier to

see a pair of electrons in a normal carbon–leaving group σ bond

than it is to recognize the dashes representing partial

bonds as a leaving group.

21.7 Additional Problems

PROBLEM 21.27 Here comes a problem for rabbits. Provide a

mechanism for this simple change.

BrCH

1. H

O, (CH

)

CHOH

2. KOH/H

(neutralize)

21.6 Summary

New Concepts

One concept dominates this chapter. The chemistry of many

organic molecules containing leaving groups is strongly influ-

enced by the presence of an internal nucleophile, a neighboring

group. Internal nucleophiles can increase the rate of ionization

of such compounds by providing anchimeric assistance in the

ionization step. Moreover, the structures of the products derived

from the ion produced through intramolecular displacement

can be subtly different from those expected of simple intermol-

ecular displacement. The three clues to the operation of a

neighboring group effect are (1) an unusual stereochemical

result, generally retention of configuration where inversion

might have been expected; (2) a rearrangement, which is

usually the result of formation of a cyclic ion through intramol-

ecular displacement of the leaving group; and (3) an unexpect-

edly fast rate.

The cyclic ions formed by neighboring group displacement

are of two kinds.There are classical species in which all bonds

are two-electron bonds, and there are more complex, nonclassi-

cal or delocalized species in which three-center, two-electron

bonding is the rule. Figure 21.40 illustrates the difference.