γ-Silyl-stabilized tertiary ions? Solvolysis of 4-(trimethylsilyl)-2-chloro-2-methylbutane

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY

J. Phys. Org. Chem. 12, 564–576 (1999)

g-Silyl-stabilized tertiary ions? Solvolysis of

4-(trimethylsilyl)-2-chloro-2-methylbutane

Leon J. Tilley¹* and V. J. Shiner, Jr²

¹Department of Chemistry, Stonehill College, 320 Washington Street, Easton, Massachusetts 02357, USA

²Department of Chemistry, Indiana University, Bloomington, Indiana 47405, USA

Received 2 November 1998; revised 10 February 1999; accepted 26 February 1999

ABSTRACT: Rate constant, isotope-effect, and product studies of the solvolysis of 4-(trimethylsilyl)-2-chloro-2-

methylbutane, 11, and its carbon analog, 2-chloro-2,5,5-trimethylhexane, 10, in aqueous ethanol and aqueous 2,2,2-

triﬂuoroethanol (TFE) indicate very little participation of the g-silyl substituent. These results are in sharp contrast to

earlier reports on secondary g-silyl substituted systems, in which the back lobe of the silicon–carbon bond has been

shown to overlap with the carbocation p-orbital to form a so-called ‘percaudally’ stabilized intermediate. While the

solvolytic behaviors of 11 and 10 are nearly identical in ethanol, differences in the TFE lead to the conclusion that

there is a minor amount of participation by the silyl substituent in that solvent. Interestingly, this observation lends

credence to an earlier suggestion that TFE is better than ethanol at stabilizing more highly delocalized ions. Copyright

1999 John Wiley & Sons, Ltd.

KEYWORDS: g-silyl stabilized tertiary ions; solvolysis; 4-(trimethylsilyl)-2-chloro-2-methylbutane

INTRODUCTION

Silicon is well known to inﬂuence reactivity at carboca-

tion centers (for reviews of silyl-substituted carbocations,

see Ref. 1). However, the exact nature and magnitude of

1

the silicon effect is highly dependent upon the position of

the silicon atom relative to the developing positive charge

and the structure and conformation of the reacting

substrate.

roughly 4 kcal mol

(1 kcal = 4.184 kJ) less effective

than an alkyl-substituted system in stabilizing carboca-

tions in solution. The decreased ability of an a-silyl

substituent to stabilize positive charge is believed to

result from poorer hyperconjugation of Si—C bonds with

the carbocation p-orbital.^1,2,6Although one gas-phase

study⁷suggests otherwise, several theoretical calcula-

tions have supported this supposition, and suggest that

the order of stability for a-substituted carbocations is H

< Si < C.^3,8

In contrast, b-silyl substituents have been shown to

stabilize carbocations greatly. Since an early report by

Ushakov and Itenberg⁹in 1937, there have been

numerous investigations to determine the magnitude

and origins of this effect. In a solvolytic study of the

conformationally restricted systems 3 and 4, Lambert et

Notwithstanding complications resulting from steric

and ground-state effects, results indicate that an a-silyl

substituent seems to stabilize carbocations relative to

hydrogen, but retards solvolysis rates relative to alkyl

substituents. Cartlege and Jones²found 2-bromo-2-

(trimethylsilyl)propane to solvolyze 38000 times slower

than the carbon analog, 2-bromo-2,3,3-trimethylbutane.

Apeloig and co-workers^3,4found 2-(trimethylsilyl)-2-

adamantyl p-nitrobenzoate to solvolyze at roughly the

same rate as 2-methyl-2-adamantyl p-nitrobenzoate, but

concluded that the destabilizing inﬂuence of the a-silyl

effect was masked by a leaving group-dependent

electronic geminal interaction which raised the ground-

state energy for the silyl-substituted nitrobenzoate. In a

study designed to minimize such complexities, Shimizu

et al.⁵determined solvolytic rates for 1 and its carbon

analog 2, and concluded that the a-silyl substituent was

*Correspondence to: L. J. Tilley, Department of Chemistry, Stonehill

College, 320 Washington Street, Easton, Massachusetts 02357, USA.

E-mail: ltilley@stonehill.edu

al.¹⁰reported rate accelerations of 2.4 Â 10¹²and

4.0 Â 10⁴, respectively, over cyclohexyl triﬂuoroacetate.

Contract/grant sponsor: Indiana University.

CCC 0894–3230/99/070564–13 $17.50

g-SILYL-STABILIZED TERTIARY IONS

565

In addition, it was possible to resolve the inﬂuence of the

b-silicon into contributions from an inductive effect and a

conformationally dependent effect believed to arise

either from hyperconjugation or from a true, bridged

ion containing a ﬁve-coordinate silicon. The conforma-

tional effect is, unless forbidden by geometry, believed to

provide the majority of the stabilizing inﬂuence of the

silyl substituent. Although some ambiguity still exists

regarding the extent of bridging in the stabilized ion,^8,11

measurement of a-deuterium isotope effects by Flem-

ing,¹²Lambert et al.¹³and Shimizu^1atend to support

hyperconjugation rather than a bridged intermediate.

The ability of a g-silyl substituent to stabilize

carbocations is an interesting phenomenon. In an early

paper, unusual reactivity of a g-silyl system was

demonstrated by Sommer et al.,¹⁴who found (3-

chloropropyl)trichlorosilane to react with ethanolic

KOH within 1 h, whereas n-hexyl chloride did not.

Indeed, it was later shown that 1,3-elimination of a g-silyl

substituent could be used to synthesize cyclopropanes.¹⁵

Fleming and co-workers^16–19reported the ability of g-

silyl substituents to control carbocation rearrangements

under Lewis acid-catalyzed conditions. During a study of

cyclopropane formation from 1,3-deoxystannylation of

norbornyl mesylates, Davis and Johnson²⁰proposed an

alternative mechanism to concerted elimination, which

involved a ‘percaudally’ stabilized intermediate wherein

the back lobe of the carbon–tin sigma bond overlapped

with the p-orbital of the carbocation to stabilize the

positive charge. In an attempt to detect such an

intermediate, g-silyl and g-stannyl substituted sulfonates

were solvolyzed.²¹In aqueous acetic acid, 4-(trimethyl-

silyl)-2-butyl methanesulfonate was found to solvolyze

7.8 times faster than 2-butyl methanesulfonate, but no

methylcyclopropane was found. It was concluded that no

g-silyl stabilized ion existed.

Figure 1. Carbocation stabilized by `Percaudal' overlap of g-

silyl substituent

(Trimethylsilyl)-2-butyl brosylate, 5, was found to react

120 times faster in 97T than the carbon analog 5,5-

dimethyl-2-hexyl brosylate, 6.²⁶Reduced b-d₃and a-d

isotope effects and unusually small or inverse b-d₂

isotope effects were also seen, as well as the formation of

signiﬁcant amounts of methylcyclopropane. Conse-

quently, the proposed mechanism for this reaction also

involved silicon-promoted carbon participation. Addi-

tionally, since the products were found to be racemic, it

was suggested that, unlike the cyclohexyl system wherein

participation occurs exclusively through a W conforma-

tion, the g-silyl substituent can stabilize the transition

state for the straight-chain system by either a W or an

endo-sickle conformation. Several later studies of

stereospeciﬁcally labeled straight-chain g-silyl-substi-

tuted secondary brosylates lent further support to this

conclusion.^27,28

While there is ample evidence for participation

involving a g-silyl substituent for secondary systems,

surprisingly little work has been done on the correspond-

ing tertiary systems. Fleming and co-workers^16–19

reported Lewis acid-catalyzed rearrangements for g-

silyl-substituted tertiary alcohols. Grob and co-work-

ers^29,30measured solvolysis rates for the substituted

adamantyl bromides 7 and 8 in 80% (v/v) aqueous

Shiner and Ensinger^22,23have described a number of g-

silyl-substituted secondary systems whose solvolyses

have conclusively been demonstrated to involve per-

caudally stabilized ions. During the solvolyses of cis- and

trans- 3-(trimethylsilyl)cyclohexyl brosylates (brosy-

late = 4-bromobenzenesulfonate), the cis- (but not the

trans-) isomer was found to react two orders of

magnitude faster than carbon analogs in 97% aqueous

2,2,2-triﬂuoroethanol (97T). This rate increase, and also

the difference in reactivity between the cis and trans

systems, was indicative of a conformationally dependent

g-silicon effect. A b-d₄isotope effect of nearly unity, the

presence of a bicyclohexane 1,3-elimination product and

an ethanol–TFE plot²⁴markedly different from the other

systems provided further evidence of the involvement of

percaudally stabilized ion in a ‘W’ conformation, as

shown in Figure 1. Theoretical calculations are also

consistent with an orientation-dependent g-silicon ef-

fect.²⁵

ethanol (80E) and found only modest rate accelerations

of 8.6-and 33-fold, respectively. Grob and Waldner^31,32

also studied the ethanolysis of the tertiary stannyl

chloride, 9, and its carbon analog, 2-chloro-2,5,5-

The g-silyl effect has also been evidenced in

conformationally unrestricted straight-chain systems. 4-

trimethylhexane, 10. Product studies of 9 indicated the

J. Phys. Org. Chem. 12, 564–576 (1999)

566

L. J. TILLEY AND V. J. SHINER, JR

Table 1. Rate constants for solvolyses of tertiary chlorides at 25°C

Solvent

5

Compound

97T (k Â 10 s)

97T (k_rel)

80E (k Â 10⁵s)

80E (k_rel)

130.9 (0.2)

2.35

1.185 (0.003)

1.13

55.75 (0.02)

13.19^a

1.00

1.050 (ꢀ0)

1.00

0.237

0.9338^b

0.889

a

Ref. 34.

^bRef. 35.

exclusive formation of dimethylcyclopropane; this com-

pound was seen to react 12–14 times faster in 80E

(calculated from rate constants reported at 50 and 60°C)

than the carbon analog. By comparison of Grunwald–

Winstein m values,³³it was concluded that this reaction

was not concerted, but involved a carbocation inter-

mediate.

Accordingly, we decided to study the solvolysis of 4-

(trimethylsilyl)-2-chloro-2-methylbutane, 11. The study

of this system was expected to provide further insight into

the nature of the stabilizing inﬂuence of a g-silyl

substituent, and to determine its effect, if any upon a

tertiary system. Measurements of rate constants, kinetic

isotope effects and product studies were made on this

system and on the carbon analog, 10.

and deuterated 97T (97% 2,2,2-triﬂurorethanol-d₃–3%

D₂O by weight). The identities and relative proportions

1

of the products were determined by H NMR and are

given in Table 3. In addition to the expected elimination

and substitution products, a small amount of the 1,2-

elimination product, 20, was also observed in 97T. This

compound results from a hydride shift to form a b-silyl-

stabilized carbocation, followed by elimination of the

trimethylsilyl group. Such a rearrangement has been

previously observed by Fleming and co-workers.^18,19

DISCUSSION

At ﬁrst glance, stabilization of the carbocation by

percaudal participation of the g-silyl substituent appears

to be absent. The results appear to conﬁrm expectations

that a tertiary carbocation intermediate exhibits a smaller

electron demand on a silyl substituent. For example, Li

and Stone³⁷have previously demonstrated in the gas

phase that the magnitude of the b-silyl effect decreases by

10 kcal mol ¹with each replacement of a b-hydrogen by

a methyl group. Consequently, it seems likely that the g-

effect would also be much reduced for the tertiary as

opposed to the secondary system.

In contrast to the secondary g-silyl brosylate, 5, which

reacted 129 times faster²⁶than the carbon analog, 6, in

97T and 2.7 times faster in 80E, the tertiary chloride 11

solvolyzed only 2.35 times as fast as its carbon analog in

97T; the rate constants are essentially identical in 80E. It

is interesting that, of the two solvents, the only

acceleration detected for the silyl compound is seen in

RESULTS

The conductometric rate constants for the solvolysis of

11 and several carbon analogs^34,35in aqueous ethanol and

2,2,2-triﬂuoroethanol at 25°C are given in Table 1. In

cases where multiple determinations of a rate constant

were performed, values in parentheses indicate average

deviations. To aid in making comparison, relative rate

constants are also indicated.

b-Deuterium kinetic isotope effects for 10 and 11 are

given in Table 2. For comparison, the isotope effects for a

number of other tertiary chlorides are also included.^35,36

In order to determine reaction products, 11 was

allowed to solvolyze for at least 10 half-lives at 25°C

in deuterated 80E (80% ethanol-d₆–20% D₂O by volume)

J. Phys. Org. Chem. 12, 564–576 (1999)

g-SILYL-STABILIZED TERTIARY IONS

Table 2. Isotope effects for solvolyses of tertiary chlorides at 25°C^a

567

Solvent

Compound

Parameter

97T

80E

b d₆(k_H/k₂

)

1.610

1.269

1.159

1.727

1.314

1.320

methyl d₃

1,1,1

^d31/2

b d₃(k_H/k₂

)

methyl d₃

b d₂(k_H/k_3,3

1,1,1 d3

)

d₂

b d₆(k_H/k₂

b d3 (k_H/k₂

)

1.761

1.327

1.452^b

1.741

1.320

1.409^b

methyl d₃

1,1,1

^d31/2

)

methyl d₃

1,1,1 d3

b d₂(k_H/k_3,3

)

d₂

1/3

b d₃(k_H/k_b

)

1.378^c

1.349^c

d₉

b d₃(k_H/k_1,1,1

b d₂(k_H/k_3,3

)

—

1.34^d

1.40^d

d₃

d₂

b d₃(k_H/k_1,1,1

b d₂(k_H/k_3,3

)

—

1.34^d

1.47^d

d₃

d₂

b d₃(k_H/k_1,1,1

b d₂(k_H/k_3,3

)

—

1.40^d

1.08^d

d₃

d₂

a

Except where noted, the isotopically labeled compounds used were at least 97% deuterated.

^bCompound was approximately 90% deuterated.

c

Ref. 35.

^dRef. 36.

97T. This observation is in accord with data for

secondary systems which also show (but to a much

greater extent) larger rate enhancements in TFE than in

ethanol.^22,23,26,28Thus, while rate studies seem to

indicate greatly diminished participation of the g-silyl

substituent, the possibility of some stabilization in 97T

still exists.

Indeed, the b d₂and b d₃isotope effects for 11 in

97T, which are signiﬁcantly smaller than those seen for

10, lend further support to the possibility of the existence

of a percaudally stabilized ion in that solvent. The W or

endo-sickle conformation required for percaudal par-

ticipation of a g-silyl substituent has been shown to

lead to greatly reduced or even inverse b d₂isotope

effects in secondary straight-chain and cyclohexyl

systems.^22,23,26,28In such instances, the dihedral angle

between the C—D bonds and the developing p-orbital of

the reacting carbon center is such as to allow for very

little hyperconjugation (for a discussion of the conforma-

tional dependence of the b-deuterium isotope effect, see

Refs^36,38), and often only the inductive effect (which

leads to inverse isotope effects) is observed. Since a

stabilized intermediate in the solvolysis of 11 would be

expected to proceed through a similar conformation (Fig.

1), the b -d₂isotope effect for this system would also be

expected to be small. Indeed, in 97T, the b -d₂effect for

11 of 1.159, although not as small as for the secondary

systems, is considerably smaller than that of 1.452 for 10.

It might be argued that the reduced b-deuterium

isotope effect is the result of a particular conformer being

sterically favored. This has been observed for 2-chloro-

2,4,4-trimethylpentane, 15 (b -d₂isotope effect 1.08).³⁶

However, conformational restriction due to steric repul-

sion between adjacent alkyl groups would not be

expected to be as important for 11. Indeed, the carbon

analog 10, which would in fact be expected to be more

congested, exhibits a typical b -d₂isotope effect of 1.452.

Furthermore, the b -d₃isotope effect for 11 in 97T (in

which the b-methyl group is freely rotating and

unaffected by conformational restrictions) is also seen

to be reduced relative to the b -d₃effect for 10. The fact

that the isotope effects are reduced to a greater extent in

97T than 80E also rules out the steric explanation.

The isotope effects also appear to reﬂect differing

J. Phys. Org. Chem. 12, 564–576 (1999)

568

L. J. TILLEY AND V. J. SHINER, JR

Table 3. Solvolysis products of 4-(trimethylsilyl)-2-chloro-2-

methylbutane

1,3-elimination products derived from loss of the

trimethylsilyl group have been observed.^{22,23,26–28}Strong

percaudal participation in the solvolysis of 11 should

therefore be accompanied by the formation of 1,1-

dimethylcyclopropane. It is therefore possible that, while

the degree of positive charge delocalization at silicon is

insufﬁcient to favor attack of a solvent molecule at

silicon to cause 1,3-elimination, the conformation

required by participation apparently tends to disfavor

rate-determining b -CH₂elimination in 97T.

Solvent

Product^a

Deuterated 97T Deuterated 80E

14.6

39.6

Using the results of this study and by drawing an

analogy with solvolytic studies of tert-butyl chlor-

ide,^34,36,39we have postulated a reasonable mechanistic

picture for the solvolyses of 10 and 11, as shown in Fig. 2.

The rate-determining step for all of the systems is

probably the formation of the solvent-separated ion pair.

However, whereas competing rate-determining elimina-

tion appears to be occurring for 10 in 97T, this does not

seem to be the case for 11, whose b -d₂isotope effect is

much smaller in that solvent. Indeed, the ion formed

during the solvolysis of 11 may gain some additional

stability from participation (albeit slight) of the g-silyl

substituent. In 80E, although a very small extent of

participation may be indicated by the slightly smaller b -

d₂isotope effect of 11 (1.32 versus 1.40), both

compounds probably react by similar mechanisms. In

neither system is evidence of competing rate-determining

elimination conclusive.

50.1

26.0

Total substitution products

64.7

21.0

65.6

20.7

12.7

13.6

1.7

0

Total elimination products

35.4

34.3

CONCLUSION

a

Expressed as mol% of total mixture.

The results of this study indicate a much smaller extent of

percaudal participation by a g-silyl substituent during the

solvolysis of a tertiary substrate than for a secondary

substrate. Considering the inherent stability of tertiary

carbocations, it is not surprising that there is a reduced

demand on the ability of the silyl substituent to delocalize

the positive charge.

amounts of competing rate-determining elimination,

depending upon the extent of participation of the silicon.

An examination of the products from the solvolysis of 11

supports this observation. The relative amounts of

elimination in 97T and in 80E are remarkably similar.

This is in sharp contrast with results obtained for tert-

butyl chloride,^34,36which shows considerably larger

amounts of elimination and also a larger b -d₃isotope

effect in TFE as compared with ethanol. The results for

the tert-butyl system have been attributed to partial rate-

determining elimination from the tight ion-pair in TFE,

with the larger b-deuterium isotope effect arising from

the contribution of a primary isotope effect.³⁶Indeed,

results similar to those obtained for tert-butyl chloride are

seen for 10, whose b -d₃isotope effect remains

essentially identical in the two solvents, but whose b-d₂

isotope effect is larger in 97T.

In spite of the evidence of participation provided by the

similarity in the amounts of elimination products in both

solvents, the absence of signiﬁcant amounts of 1,1-

dimethylcyclopropane seems to preclude the possibility

of an intermediate carbocation stabilized by strong

silicon-promoted carbon participation. For secondary g-

silyl-substituted systems, signiﬁcant amounts of cyclic

An interesting comparison can be made between our

results and Grob and co-workers’ earlier study^29,30

involving ethanolysis of the g-silyl substituted adamantyl

system 7. The rate acceleration of 8.6-fold found in 80E

is larger than our value of 1.13-fold (essentially none).

This observation is in good agreement with the expecta-

tion that percaudal participation of the silyl substituent

would be much less for the open-chain system than the

adamantyl system in which silyl substituent is held in the

conformation required for maximum participation.

Additionally, while we did not observe appreciable

rate acceleration for the open-chain silyl system 11 in

80E, Waldner and Grob³¹reported a rate acceleration of

12–14-fold for the corresponding stannyl analog 9, and

also the formation of 1,1-dimethylcyclopropane. These

results concur with the expectation that percaudal

participation is more important for a g-stannyl as opposed

to a g-silyl system, because of the greater electron-

J. Phys. Org. Chem. 12, 564–576 (1999)

g-SILYL-STABILIZED TERTIARY IONS

569

Figure 2. Proposed mechanisms for solvolysis of 4-trimethylsilyl-2-chloro-2-methylbutane and 2,5,5-trimethyl-2-chloropentane

donating ability of tin. While deuterium kinetic isotope

effects for 9 have not been measured, it is likely that they

would provide evidence for a mechanism involving

percaudal participation in the open-chain stannyl system.

Moreover, if measured, the rate acceleration in 97T might

be expected to be much greater than that in 80E.

In ethanolic solvents, the apparent near total lack of

participation of the g-silyl substituent during the

solvolysis of 11 implies that the percaudally stabilized

ion is not appreciably more stable than the open tertiary

ion. However, in TFE, there is some bias toward a

percaudally delocalized ion. Interestingly, this observa-

tion is in good accord with an earlier suggestion of

Stoelting and Shiner⁴⁰that TFE is relatively more

effective than ethanol at stabilizing larger, more

delocalized carbocations. Indeed, in view of this

evidence, it appears that the g-silyl effect is greatly

reduced for tertiary systems, but can manifest itself to

some extent under the appropriate circumstances.

EXPERIMENTAL

General

NMR spectra were recorded on a 300 MHz Varian XL-

300, a 500 MHz Bruker AM-500 or a 500 MHz Varian

^UnityINOVA 500 spectrometer. IR spectra were taken on a

Mattson Instruments 4020 Galaxy Series Fourier trans-

form IR spectrometer. Analytical gas chromatography

(GC) was carried out using a Hewlett-Packard Model

5890 gas chromatograph, equipped with a 50 m Â 0.2mm

i.d. (0.33 mm ﬁlm thickness) HP-5 column and a ﬂame

ionization detector. Preparative GC was performed using

a Varian Aerograph Series 2700 gas chromatograph,

equipped with a 6 ft Â 1/4 in column of 20% OV101 on

Chromosorb P (60–80 mesh) and a thermal conductivity

detector. Mass spectra were recorded on a Kratos MS80

RFAQQ instrument. GC–mass spectra were obtained on

a Hewlett-Packard GC/MSD 5971 instrument, equipped

J. Phys. Org. Chem. 12, 564–576 (1999)

570

L. J. TILLEY AND V. J. SHINER, JR

with a 60 m Â 0.25mm i.d. SPB-5 column. Melting-

points and boiling-points are uncorrected. Boiling-points

for molecular distillations were not directly determined

since the still did not have provisions for a thermometer.

0.056mmol. The tube was sealed and the mixture was

allowed to react at 25°C for at least 10 half-lives.

Relative amounts of the products were determined from

integrated peak areas of the 500 MHz ¹H NMR spectrum;

the error using this technique is estimated to be 2–3%.

Peaks unique to each product are as follows: 16a,

1.475 ppm (m, CH₂, 3-carbon, b- to Si); 17a, 1.502 (m,

CH₂, 3-carbon, b- to Si); 18, 1.393 (d, CH₂, 4-carbon, a-

to Si, J = 8.5 Hz), 1.56 (s with ﬁne splitting, allylic CH₃),

1.68 (s with ﬁne splitting, allylic CH₃), 5.213 (triplet of

septets, vinyl H, J = 8.5 Hz, 1.4 Hz); 19, 0.656 (m, CH₂,

4-carbon, a- to Si), 1.72 (s, allylic CH₃), 2.027 (m, CH₂,

3-carbon, b- to Si), 4.66 (m, vinyl H), 4.72 (m, vinyl H);

20, 0.976 (d, 2CH₃’s), 2.25 (m, CH), 4.85 (overlapped d

of d, E-terminal vinyl H, J_cis= 10.4 Hz), 4.96 (overlapped

d of d of d; looks like d of t, Z-terminal vinyl H,

J_trans= 17.3 Hz, J_gem= 1.7 Hz, J_allylic= 1.7 Hz), 5.83

(overlapped d of d of d, internal vinyl H). (Alcohol 16a

and ether 17a were distinguished by spiking the reaction

mixture with known 16. The multiplets used to determine

the 16a/17a ratio were of similar shape but partially

overlapped. An estimate of the relative amounts of 16a

and 17a was obtained by measuring and comparing the

relative heights of the outermost peak of each multiplet.)

Materials

Nitrogen. In order to ensure removal of carbon dioxide

and water, all nitrogen used for transferring and distilling

conductivity solvents was puriﬁed by passage through a 3

˚

ft glass column packed with 4 A molecular sieves.

Conductivity solvents. Conductivity water and con-

ductivity ethanol were prepared as described by Murr⁴¹

and Tilley.⁴²Conductivity 2,2,2-triﬂuoroethanol was

prepared as described by Shiner et al.³⁴and Tilley.⁴²

Kinetics

All rate constants for this work were determined

conductometrically at 25°C using a bipolar pulsed

conductance (BIPCON) instrument, based on the design

of Caserta et al.⁴³with the modiﬁcations of Ensinger⁴⁴

and Tilley.⁴²The constant-temperature oil-bath em-

ployed for kinetic measurements was designed by

Murr,⁴¹with computer-regulated temperature control

later added using modiﬁcations of Ensman,⁴⁵Withnell,⁴⁶

Wilgis⁴⁷and Tilley.⁴²The conductance cells used were

made in this laboratory, based on the design of Murr,⁴¹

using modiﬁcations by Rapp,⁴⁸Tomasik,⁴⁹Wilgis⁴⁷and

Tilley.⁴²Data calibration and acquisition were performed

on a JCC Systems computer containing an Intel 33 MHz

386 processor with a 387 math co-processor chip, using

software described by Tilley,⁴²which is based on earlier

programs of Ensinger,⁴⁴Wilgis,⁴⁷Tomasik,⁴⁹Sporle-

der⁵⁰and Stoelting.⁵¹Calibration was done with the aid

of a Dial-An-Ohm resistor box (General Resistance,

North Branford, CT, USA). After acquisition, the raw

data were then converted into resistance–time data using

a program written by Tilley⁴²and analyzed using a non-

linear, doubly weighted least-squares program originally

written by Murr,⁴¹with modiﬁcations by Buddenbaum,⁵²

Vogel,⁵³Pinnick,⁵⁴Bowersox,⁵⁵Tomasik,⁴⁹Wilgis,⁴⁷

Stoelting,⁵⁶Ensinger⁵⁷and Tilley.⁴²

Product study in deuterated 80E. The above procedure

was repeated by adding 10 ml, 0.049mmol of 11 and

6.5 ml, 0.056mmol of 2,6-lutidine to a tube containing

ethanol-d₆(Cambridge Isotope Laboratories), 800 ml, and

D₂O, 200 ml. Peaks unique to each product are as follows:

16a, 0.496 ppm (m, CH₂, 4-carbon, a- to Si), 1.136 (s,

geminal CH₃s); 17b, 0.450 (m, CH₂, 4-carbon, a- to Si);

18, 1.342 (d, CH₂, 4-carbon, a- to Si, J = 8.5 Hz), 1.5 (s

with ﬁne splitting, allylic CH₃), 1.646 (s with ﬁne

splitting, allylic CH₃), 5.103 (triplet of septets, vinyl H);

19, 0.6113 (m, CH₂, 4-carbon, a- to Si), 1.679 (s with ﬁne

splitting, allylic CH₃), 1.962 (m, CH₂, 2-carbon, b- to Si),

4.615 (m, vinyl H), 4.648 (m, vinyl H).

Synthetic Procedures

Synthesis of g-silyl-substituted chlorides. (Trimethyl-

silyl)methyl iodide, 21. This was synthesized as de-

scribed by Whitmore and Sommer.⁵⁹Under nitrogen,

approximately 1500ml of acetone were puriﬁed by

reﬂuxing for 6 h over CaO and KMnO₄, followed by

fractionation through a 40 cm Vigreaux column (b.p.

56°C; lit.⁶⁰b.p. 56.2°C).

NMR product studies of 4-(trimethylsilyl)-2-

chloro-2-methylbutane, 11

To sodium iodide, 215.1 g, 1.435 mol, dissolved in

approximately 1200ml of acetone, was added (trimethyl-

silyl)methyl chloride, 101.3 g, 0.8258 mol, all at once,

using an additional 200ml of acetone for rinsing. The

mixture was stirred under nitrogen at reﬂux for 24 h.

Almost immediately, a white precipitate of NaCl formed

that thickened over time, eventually hindering agitation.

The cooled solution was ﬁltered through a medium-

Product study in deuterated 97T. In an NMR tube with

a Pyrex extension was placed 2,2,2-triﬂuoroethanol-d₃

(Cambridge Isotope Laboratories), 706.5 ml, 970mg, and

D₂O, 30 ml. To this was added 4-(trimethylsilyl)-2-

chloro-2-methylbutane, 11, 10 ml, 0.049mmol [density

1

0.869gml

(Ref. 58)], and 2,6-lutidine, 6.5 ml,

J. Phys. Org. Chem. 12, 564–576 (1999)

g-SILYL-STABILIZED TERTIARY IONS

571

porosity, 600ml fritted funnel, rinsing with 200ml of

wash acetone. Acetone was then removed by fractiona-

tion through a 40 cm Vigreaux column. After cooling,

pentane (50ml) and water (250ml) were poured down the

Vigreaux column into the pot. The mixture was

transferred into a separating funnel, using 150ml of

water and 250ml of pentane, shaken and separated. The

aqueous layer was extracted with pentane (6 Â 50mL).

The combined organic layers were washed with 10%

aqueous sodium thiosulfate (3 Â 100ml) to remove free

iodine, and then with water (2 Â 100ml). The pentane

solution was dried (MgSO₄), ﬁltered, and concentrated by

rotary evaporation in a room-temperature bath.

followed by distillation from CaO. Methyl acetoacetate

was puriﬁed by distillation at reduced pressure (b.p. 69°C

at 13mmHg). To a suspension of 0.80 g, 0.10 mol, of LiH

in 100ml of DMF were added dropwise, with stirring,

11.6 g, 0.10 mol of methyl acetoacetate during 10 min.

Hydrogen evolved slowly, requiring stirring for an

1

additional 3 h. H NMR of an aliquot (500 MHz, neat)

showed quantitative formation of the lithium enolate.

Next, 21, 21.4 g, 0.10 mol, was added dropwise over

10 min. The mixture was stirred for 20 h at room

temperature, 48 h at 60°C and 24 h at 100°C to ensure

total reaction.

The cooled reaction mixture was poured into 1 l of

water, with additional rinses of 200 and 300ml. The

combined pentane extracts (5 Â 200ml) were washed

with water (3 Â 500ml), dried (MgSO₄) and solvent was

removed by rotary evaporation. A distillation fraction

(113–122°C at 20mmHg) was found by NMR (CDCl₃,

500 MHz) to contain mostly the desired product. This,

when re-distilled (10 cm Vigreaux column; lit.⁶¹b.p.

60°C at 2.0mmHg) was sufﬁciently pure by NMR for use

in the next reaction. A total of 8.76 g of 23 were

synthesized (43%). NMR indicated both a keto and an

The residue was then distilled (10 cm Vigreaux

column, b.p. 137–140°C; lit.⁵⁹b.p. 139.5°C) to give

137.8 g of 21 (78%): ¹H NMR (CDCl₃, 300 MHz) ꢀ 0.147

(s), 2.00 (s); ¹³C NMR (CDCl₃, 75 MHz), ꢀ 12.06, 1.59.

Ethyl 3-(trimethylsilyl)propionate, 22. Compound 22

was synthesized using a procedure by Sommer and

Marans.⁵⁸Under nitrogen, sodium metal, 6.51 g, 0.2831

mol, was dissolved in 100ml of commercial anhydrous

ethanol, completion requiring applied heat (reﬂux).

Next, puriﬁed ethyl acetoacetate,⁶⁰33.5 g, 0.2574 mol,

was added at reﬂux during 15 min. Then 21, 50.0 g, 0.234

mol, was added over 15 min.

1

enol form. H NMR (CDCl₃, 500MHz, keto form), ꢀ

0.009 (s, 9H), 1.028 (m, 1H), 1.150 (m, 1H), 2.207 (s,

1

3H), 3.425 (m, 1H), 3.714 (s, 3H); H NMR (CDCl₃,

After 36 h at moderately fast reﬂux, the cooled mixture

was stirred with 6.76 g, 0.049 mol, of sodium hydro-

gensulfate monohydrate, for 10 min (insolubles, pH

basic), heated (clariﬁed) and gently reﬂuxed for 30–

40 min (brownish precipitate cloudiness, pH 5–6). The

mixture was cooled, transferred and agitated with 500ml

of diethyl ether and ultimately ﬁltered on a ﬁne-porosity

frit to remove precipitated salts (NaI, NaHSO₄, Na₂SO₄).

Most of the ether was removed by distillation. The

concentrate was transferred to a 250ml round-bottomed

ﬂask. The remaining ethanol and ethyl acetate were

mainly removed by distillation through a 40 cm Vigreaux

column, whereupon NaI precipitation occurred. The

cooled residue was transferred with 400ml of pentane,

causing further precipitation of NaI. The ﬁltrate was

washed with water (3 Â 100ml), dried (CaSO₄) and

ﬁltered through a pentane-slurried 4 cm bed of alumina

gel (Fisher, basic, Brockman I) in a medium-porosity,

350ml fritted funnel, rinsing the bed with 200ml

additional pentane. Pentane was mainly removed by

distillation through a 40 cm Vigreaux column, the last

traces by blowing with a gentle nitrogen stream, giving

500MHz, enol form), ꢀ 0.033 (s, 9H), 1.530 (s, 2H),

1.942 (s, 3H), 3.715 (s, presumably 3H; very close to

larger peak of keto form), 12.5 (s); ¹³C NMR (CDCl₃,

125 MHz, keto form), ꢀ 1.420, 15.21, 27.58, 52.05, 55.30,

171.17, 202.28; ¹³C NMR (CDCl₃, 125 MHz, enol form),

ꢀ

1.13, 14.52, 19.11, 51.14, 97.10, 169.64, 173.60; IR

1

(neat), 1744.9 cm (s, C

form), 1719.3 cm ¹(s, C

form), 1648 (m, C O str for conjugated C

form), 1613.7 (m, C

m/z 203.1 (M 1, 1.3%), 202.1 (M , 2.1%), 187

(43.4%), 170 (6.0%), 160 (12.8%), 159 (24.5%), 155

(27.3%), 143 (24.2%), 127 (18.9%), 113 (27.6%), 89

(58.7%), 75 (20.5%), 73 (100.0%), 59 (17.8%), 55

(66.3%), 45 (12.8%), 43 (29.3%).

=O str for ester C=O of keto

=O str for ketone C=O of keto

=

=O of enol

=

C str of enol form); MS (CI, NH₃),

Methanol-d, 24^63,64. Methanol-d was synthesized ac-

cording to the procedure described by Streitwieser et al.

⁶⁴Dimethyl sulfate was puriﬁed by distillation at reduced

pressure (b.p. 81–83°C at 15mmHg). Dimethyl carbo-

˚

nate, treated with 4 A molecular sieves for 4 days, was

fractionally distilled through a 50 cm vacuum-jacketed

glass column packed with glass helices (b.p. 91°C;C).

To dimethyl carbonate, 414 g, 4.60 mol, and D₂O,

102.6 g, 5.12 mol, in a 1 l one-necked round-bottomed

ﬂask, dimethyl sulfate, 16 g, 0.13 mol, was added; a stir

bar was inserted and a 60 cm reﬂux column afﬁxed,

protected by Drierite (Hammond). After a 120 h reﬂux,

¹H NMR (neat, 500 MHz) showed only 0.3 mol% of

dimethyl carbonate remaining. After cooling, the reﬂux

column was replaced by a 50 cm vacuum-jacketed

1

20.39 g of 22 (50%): H NMR (CDCl₃, 500 MHz) ꢀ

0.035 (s, 9H), 0.799 (m, 2H), 1.21 (t, 3H), 2.23 (m, 2H),

4.08 (q, 2H); ¹³C NMR (CDCl₃, 125 MHz), ꢀ 2.05,

11.60, 14.16, 28.85, 60.18, 175.01.

Methyl2-(trimethylsilylmethyl)acetoacetate, 23^{19, 61, 62}

.

This was prepared by a procedure based on those of

Sommer and Marans⁵⁸and Fleming and Godhill.⁶²DMF

was puriﬁed by stirring over KOH pellets for 15 min,

J. Phys. Org. Chem. 12, 564–576 (1999)

572

L. J. TILLEY AND V. J. SHINER, JR

column packed with glass helices. Methanol-d was

collected at a reﬂux ratio of 10:1 (b.p. 60.5°C at

742.6mmHg). H NMR indicated approximately 99%

deuteration. A 234.7 g amount of 24 was obtained (77%):

¹H NMR (500MHz, neat), ꢀ 3.35 (relative to added

tetramethylsilane).

4-(Trimethylsilyl)-2-methyl-2-butanol, 16⁵⁸. Under ni-

trogen, to 30ml, 0.09 mol, of 3 M methylmagnesium

bromide in diethyl ether (Aldrich) was added, with

stirring, at a rate to maintain moderate reﬂux, a solution

of 22, 5.0 g, 0.0287 mol, pre-dissolved in 30ml of

anhydrous diethyl ether from a freshly opened can. After

addition the mixture was stirred overnight at reﬂux. At

room temperature after gradual addition of 20ml of water

with agitation, the ﬂask contents were transferred to

150ml of water in a separating funnel, with the aid of

200ml more water. While maintaining pH 6–7 by added

glacial acetic acid, the aqueous layer was extracted with

diethyl ether (1 Â 100, 2 Â 75ml). The combined ether

layers were washed with saturated sodium hydrogencar-

bonate (2 Â 50ml) and water (2 Â 50ml) and dried

(CaCl₂). Diethyl ether was removed by rotary evapora-

tion (35°C bath). Molecular distillation at 4mmHg (bath

50°C) (lit.⁵⁸b.p. 48°C at 4mmHg) gave 3.17 g of 16

1

Methyl 3-(trimethylsilyl)propionate-2,2-d₂, 25^19,65

.

Compound 25 was prepared by reverse Claisen con-

densation of 23 in methanol-d, analogous to Sommer and

Marans’ preparation of 22.⁵⁸To 60ml of anhydrous

diethyl ether were added 2.2 g of a 60% dispersion of

NaH in mineral oil, 0.055 mol NaH. An additional 15ml

of diethyl ether were used to facilitate transfer of the NaH

into the ﬂask. To the suspension of NaH in diethyl ether

was added 23, 8.54 g, 0.042 mol, over a period of 15 min,

such that hydrogen evolution was not too rapid.

Methanol-d, 5ml, was then added over 5 min, such that

HD evolution was not too rapid, forming a thick,

unstirrable paste. Additional methanol-d, 55ml, added

all at once with stirring, led to dissolution to give a

slightly cloudy, pale yellow solution. This mixture was

heated, distilled through a Vigreaux column to a b.p. of

61°C;C and reﬂuxed (condenser replacing column) under

nitrogen for another 45 h.

The cooled mixture was neutralized to pH 6–7 with

2.4ml of glacial acetic acid-d (Aldrich) and poured into

250ml of water, with an additional 50ml of water for

rinsing. This solution was extracted with pentane

(4 Â 50ml). The combined organic layers were washed

with water (2 Â 60ml) and dried (CaSO₄). Pentane was

mainly distilled through a 10 cm Vigreaux column. The

residue was then distilled to give 4.53 g of colorless

liquid. A ¹H NMR spectrum showed the desired

compound, but integration of residual protons at the 2-

position showed only approximately 92% deuteration,

necessitating further exchange.

1

(69% yield): H NMR (CDCl₃, 500 MHz), ꢀ 0.018 (s,

9H), 0.485 (m, 2H), 1.18 (s, 6H), 1.38 (s, OH, variable

shift), 1.41 (m, 2H); ¹³C NMR (CDCl₃, 125 MHz), ꢀ

1.92, 10.48, 28.54, 37.79, 71.56.

4-(Trimethylsilyl)-2-methyl-d₃-2-butanol-1,1,1-d₃, 16b.

Into 85ml of 1 M methylmagnesium-d₃iodide in diethyl

ether (Aldrich), 0.85 mol, stirred under nitrogen were

added 5.00 g, 0.287 mol of 22 at a gradual rate such as to

maintain moderate reﬂux. After an additional 90 min of

stirred reﬂux, the cooled solution was treated gradually

with 30ml of saturated NH₄Cl. The ether layer was

decanted and three additional extractions were performed

(1 Â 100, 2 Â 25ml). The combined organic layers were

washed with 5% aqueous NH₄Cl (3 Â 50ml) and water

(2 Â 50ml) and dried over Na₂SO₄. Most of the diethyl

ether was removed by distillation through a 10 cm

Vigreaux column (warm water-bath) and the remainder

by rotary evaporation (room temperature bath). The

residue was puriﬁed by molecular distillation (8mmHg).

The incompletely deuterated 25, 4.52 g, 0.028 mol,

was added to a solution of metallic sodium, 0.19 g,

0.0083 mol, pre-reacted in 30ml of methanol-d. After a

48 h reﬂux, this was cooled, neutralized with 0.6ml of

glacial acetic acid-d, and treated with 150ml of water.

This was extracted with pentane (4 Â 50ml); the

combined organic layers were washed with water

(2 Â 50ml) and dried (MgSO₄). Solvent was distilled

(10 cm Vigreaux column). The residue was then distilled

on a molecular still (58mmHg, bath 90–100°C; lit.¹⁹b.p.

undeuterated 68°C at 18mmHg.) The ¹H NMR spectrum

(CDCl₃, 500 MHz) now indicated 97% deuteration. GC

indicated 93% purity, adequate for the next step. A total

1

Integration of residual methyl protons in the H NMR

spectrum showed deuterium incorporation in the desired

positions to be at least 99%. A total of 3.51 g of 16b were

1

obtained (74%): H NMR (CDCl₃, 500 MHz), ꢀ 0.045

(s, 9H), 0.457 (m, 2H), 1.38 (m, 2H), 1.61 (broad singlet,

variable, OH); ¹³C NMR (CDCl₃, 125 MHz), ꢀ 1.95,

10.40, 27.50 (septet), 37.64, 71.20.

4-(Trimethylsilyl)-2-methyl-2-butanol-3,3-d₂, 16c. This

was synthesized from 25 using the same procedure as for

16b except that 3 M methylmagnesium bromide was used

instead of 1 M methylmagnesium-d₃iodide. In this work-

up, after removing most of the diethyl ether through a

10 cm Vigreaux column, the residue was directly distilled

on a molecular still (4mmHg, bath 70–85°C), affording

1

of 3.48 g of 25 was obtained (51%): H NMR (CDCl₃,

500MHz), ꢀ 0.017 (s, 9H), 0.810 (s, 2H), 3.65 (s, 3H);

¹³C NMR (CDCl₃, 125 MHz), ꢀ 2.16, 11.27, 28.07

1

(multiplet), 51.35, 175.41; IR (neat), 3466 cm

overtone of C

str), 2130 (m, C–D str), 1739 (s, C

(m,

2.31 g of 16c. H NMR (CDCl₃, 500 MHz) indicated

=

O str), 2965 (s), 2896 (s), 2222 (m, C–D

impure alcohol; GC showed 96% purity. After prepara-

=

O str), 1435 (s), 1280

tive GC, 1.84 g of 99.99% pure 16c was obtained (54%);

1

(vs), 1204 (vs), 1107 (s), 842 (vs).

integration of the residual methylene protons in the H

J. Phys. Org. Chem. 12, 564–576 (1999)

g-SILYL-STABILIZED TERTIARY IONS

573

NMR spectrum showed the extent of deuteration in the

desired positions to be at least 97%. H NMR (CDCl₃,

outlet, with the inlet monitor connected to a lecture bottle

of ethylene, were placed 50ml of pentane and 28.1 g,

0.303 mol, of tert-butyl chloride. While stirring, the ﬂask

was cooled to 60°C in a dry-ice–acetone bath and

2.81 g, 0.021 mol, of aluminum chloride were added.

Ethylene was then bubbled through the mixture and the

dry-ice–acetone bath was removed. At 20°C, vigorous

absorption of ethylene was indicated by cessation of gas

ﬂow from the outlet bubbler, which required increasing

the gas ﬂow to prevent suck-back. A simultaneous rapid

temperature rise required prompt restoration of dry-ice

bath cooling. Cooling to about 30 to 40°C (slow

absorption) and allowing to warm to 10°C (rapid

absorption) were cycled until ethylene absorption ceased,

as seen by equal gas ﬂow in the inlet and outlet bubblers.

At 40°C, the pentane solution was decanted from the

solids, using an additional 50ml of pentane to rinse. The

pentane solution was then washed with water (2 Â 50ml),

saturated NaHCO₃, again with water (2 Â 50ml) and

dried over K₂CO₃.

1

500 MHz), ꢀ 0.023 (s, 9H), 0.463 (s, 2H), 1.175 (s, 6H),

1.46 (broad s, variable, OH); ¹³C NMR (CDCl₃,

125 MHz), ꢀ

1.91, 10.26, 28.48, 36.98 (quintet),

1

71.45; IR (neat), 3374 cm (s, broad), 2967 (s), 2908

(s), 2183 (m), 2128 (w), 2081 (w), 1364 (m), 1249 (s),

1184 (s), 863 (s), 835 (s).

4-(Trimethylsilyl)-2-chloro-2-methylbutane, 11. This

was prepared according to the method described by

Sommer and Marans.⁵⁸In a separating funnel, 1.8 g,

0.011 mol, of 16 was shaken for 15 min with 30ml of

concentrated HCl. Pentane extracts (4 Â 25ml) were

washed with 5% NaHCO₃(3 Â 30ml) and water

(2 Â 40ml) and dried (CaCl₂). Solvent rotary evaporation

followed by molecular distillation at 55mmHg (lit. b.p.⁵⁸

1

90°C at 55mm Hg) gave 0.80 g of 11 (40%): H NMR

(CDCl₃, 500 MHz), ꢀ 0.002 (s, 9H), 0.647 (m, 2H), 1.550

(s, 6H), 1.705 (m, 2H); ¹³C NMR (CDCl₃, 125 MHz), ꢀ

1.91, 11.76, 31.67, 40.49, 72.82; gated-decoupled ¹³C

NMR (CDCl₃, 125 MHz), ꢀ 1.90 (q, J = 119 Hz), 11.75

(t, J = 119 Hz), 31.66 (q, J = 128 Hz), 40.48 (t,

Pentane was removed through a 14 cm vacuum-

jacketed column packed with glass helices. The remain-

ing liquid was fractionally distilled; the large fraction

(b.p. 115–119°C; lit.⁶⁶b.p. 115°C) comprising 27.62 g of

26 (76%): ¹H NMR (CDCl₃, 500 MHz), ꢀ 0.929 (s, 9H),

1.73, (m, 2H), 3.52 (m, 2H); ¹³C NMR (CDCl₃,

125 MHz), ꢀ 29.28, 30.80, 41.59, 46.95.

J = 125 Hz), 72.85 (s); GC–MS, m/z 165 (³⁷Cl M CH₃,

1.3%), 163 (³⁵Cl M CH₃, 3.8%), 95 (27.8%), 93

(78.2%), 73 (100%), 70 (70.8%), 55 (36.6%).

4-(Trimethylsilyl)-2-chloro-2-methyl-d₃-butane-1,1,1-

d₃, 11a. Compound 11a was synthesized from 16b using

the same procedure as for 11, except that most of the

pentane was distilled out through a 10 cm Vigreaux

column with an ice–water-cooled condenser, and the last

traces by rotary evaporation from a small ﬂask at ambient

temperature. The residue was puriﬁed on a molecular still

(56mmHg). Integration of residual methyl protons in the

¹H NMR spectrum showed deuterium incorporation in

the desired positions to be approximately 99%. A total of

4,4-Dimethylpentanoic acid, 27^67,68. Under nitrogen, to

8 g, 0.33 mol, of oven-dried Mg turnings and 50ml of

anhydrous diethyl ether was added a portion of a solution

of 21.0 g, 0.174 mol, of 26 in 150ml of anhydrous diethyl

ether. After brief heating with a heat gun had initiated

reaction, the remaining chloride solution was added to

maintain reﬂux. Thereafter, a 55°C bath was applied to

continue reﬂux another 1.5 h.

The cooled solution was poured into 101.5 g of dry-ice

pellets. The resulting thick slurry required additional

diethyl ether to facilitate stirring. When most of the dry-

ice had evaporated, 250ml of 1 M HCl were added, and

stirring was continued until the dry-ice was gone. The

ether layer was separated, combined with additional ether

extracts (2 Â 100ml), dried (CaSO₄), and the ether

removed by rotary evaporation.

The residue, dissolved in 250ml of 5% aqueous NaOH,

was washed with diethyl ether (4 Â 100ml), discarding

the ether layers. The aqueous layer was acidiﬁed by

gradual addition of concentrated HCl; the product

separated as an oil. The mixture was then extracted with

CH₂Cl₂(3 Â 100ml). The combined CH₂Cl₂layers were

washed with water and dried. Rotary evaporation on a

warm water-bath gave 16.1 g of crude product, which

was vacuum distilled (b.p. 104–108°C at 13–14mmHg;

lit.⁶⁸b.p. 105°C at 13mmHg) to give 13.22 g of 27

(58%): ¹H NMR (CDCl₃, 500 MHz), ꢀ 0.907 (s, 9H), 1.56

(m, 2H), 2.32 (m, 2H), 11.32 (very broad singlet, OH).

1

2.5 g of 11a was obtained (75%): H NMR (CDCl₃,

500 MHz), ꢀ 0.00 (s, 9H), 0.642 (m, 2H), 1.697 (m, 2H);

¹³C NMR (CDCl₃, 125 MHz), ꢀ 1.90, 11.73, 30.72

(septet, J = 19 Hz), 40.36, 72.39.

4-(Trimethylsilyl)-2-chloro-2-methylbutane-3,3-d₂,

11b. Compound 11b was synthesized from 16c using the

same procedure as for 11a. Integration of the residual

1

methylene protons in the H NMR showed deuterium

incorporation in the desired positions to be not less than

1

97%. The yield was 1.12 g (55%). H NMR (CDCl₃,

500 MHz), ꢀ 0.002 (s, 9H), 0.629 (s, 2H), 1.548 (s, 6H);

¹³C NMR (CDCl₃, 125 MHz), ꢀ 1.897, 11.56, 31.61,

39.73 (quintet), 72.66.

Synthesis of carbon analogs. 1-Chloro-3,3-dimethyl-

butane, 26. The synthesis followed Schmerling’s

procedure.⁶⁶Into a three-necked, 100ml round-bottomed

ﬂask equipped with a thermometer and a bubbler-

monitored gas inlet (nebulator), and a bubbler-monitored

J. Phys. Org. Chem. 12, 564–576 (1999)

574

L. J. TILLEY AND V. J. SHINER, JR

1

Ethyl 4,4-dimethylpentanoate, 28. A 13.22 g, 0.10 mol,

amount of 27 was reﬂuxed with a catalytic amount of

sulfuric acid in 370ml of ethanol for 18 h. The mixture

was neutralized with aqueous NaHCO₃and most of the

ethanol distilled. The residue, in pentane, was washed

with aqueous NaHCO₃, then water, and dried (K₂CO₃).

Pentane was mainly distilled at ambient pressure, and the

residue under vacuum (b.p. 105–109°C at 105mmHg;

lit.⁶⁹b.p. 60–62°C at 8.0mmHg) to give 9.44 g of 28

(64%): ¹H NMR (CDCl₃, 500 MHz), ꢀ 0.87 (s, 9H), 1.23

(t, 3H), 1.52 (m, 2H), 2.25 (m, 2H), 4.10 (q, 2H).

seen in the H NMR spectrum; it was assumed that

deuterium incorporation was greater than 99%. ¹H NMR

(CDCl₃, 500 MHz), ꢀ 0.854 (s, 9H), 1.194 (m, 2H), 1.395

(m, 2H), 1.496 (s, broad, variable, OH); ¹³C NMR

(CDCl₃, 125 MHz), ꢀ 28.13 (septet), 29.29, 29.87, 38.04,

38.38, 70.68.

2,5,5-Trimethyl-2-hexanol-3,3-d₂, 29b. Compound 29b

was prepared by addition of 3 M methylmagnesuim

bromide (Aldrich) to 28a, as in the synthesis of 16.

Residual methylene protons could not be detected in the

¹H NMR spectrum; it was assumed that deuterium

incorporation was at least 97%, that which was

determined for 28a. ¹H NMR (CDCl₃, 500 MHz), ꢀ

0.865 (s, 9H), 1.184 (s, 6H), 1.197 (s, 2H), 1.342 (s,

broad, variable, OH).

Ethyl 4,4-dimethylpentanoate-2,2-d₂, 28a. This was

prepared by repeated base-catalyzed exchange of 9.44 g

of the undeuterated ester, 28, with ethanol-d–ethoxide.

Ethanol-d was removed after each exchange thus: the

mixture was neutralized with glacial acetic acid-d

(Aldrich). Most of the spent ethanol-d was removed by

distillation. The residue, in pentane, was washed with

water, dried (MgSO₄) and the solvent distilled. After 2–3

exchanges, the deuterated ester was puriﬁed by vacuum

distillation (b.p. ca 72°C at 20mmHg) to give 5.40 g of

2-Chloro-2,5,5-trimethylhexane, 10. This was synthe-

sized by a procedure of Grob and Waldner.³²Compound

29, 1.7 g, 0.117 mol, was shaken in a separating funnel

for 15 min with 33ml of concentrated HCl. This mixture

was extracted with pentane (3 Â 75ml). The combined

organic layers were washed with 5% NaHCO₃

(4 Â 50ml) and water (3 Â 50ml) and dried (CaCl₂).

Pentane was removed by rotary evaporation at ambient

temperature and the residue was distilled at 13mmHg in a

molecular still (lit.³²b.p. 46.5–47.5°C at 11mmHg) to

give 0.45 g of 10 (24%): ¹H NMR (CDCl₃, 500 MHz), ꢀ

0.894 (s, 9H), 1.355 (m, 2H), 1.567 (s, 6H), 1.706 (m,

2H); ¹³C NMR (CDCl₃, 125 MHz), ꢀ 29.30, 29.89, 32.39,

38.79, 40.90, 71.52.

1

28a (56%). H NMR analysis of the residual methylene

protons showed deuterium incorporation to have taken

place to approximately 97%. ¹H NMR (CDCl₃,

500 MHz), ꢀ 0.869 (s, 9H), 1.231 (t, 3H), 1.509 (s, 2H),

4.09 (q, 2H).

2,5,5-Trimethyl-2-hexanol, 29. This was synthesized

according to Grob and Waldner.³²To the ethereal

Grignard reagent prepared from 3.0 g, 0.123 mol, of

oven-dried magnesium turnings and 26, 10.0 g, 0.083

mol, was added dry acetone (puriﬁed as described for the

synthesis of 21), 5.89 g, 0.101 mol, dropwise so as to

maintain the ether at moderate reﬂux. After addition of

acetone, 2 M NH₄Cl was added dropwise until the pH of

the aqueous layer was neutral. The separated ether layer

was combined with ether extracts (3 Â 50ml); the

combined layers were washed with water (3 Â 50ml)

and dried (Na₂SO₄). The ether was removed by rotary

evaporation and the residue distilled at 12.5mmHg (b.p.

62–64°C; lit.³²b.p. 61.5–62°C at 10mmHg) to give

4.91 g of 29 (41% yield). Preparative GC of the crude

product, 1.78 g, afforded pure alcohol 29: ¹H NMR

(CDCl₃, 500 MHz), ꢀ 0.859 (s, 9H), 1.181 (s, 6H,

partially overlapped with multiplet at ꢀ 1.20), 1.20 (m,

2H, partially overlapped with singlet at ꢀ 1.181), 1.412

(m, 2H), 1.464 (s, broad, variable, OH); ¹³C NMR

(CDCl₃, 125 MHz), ꢀ 29.12, 29.29, 29.88, 38.09, 38.49,

70.99.

2-Chloro-2-methyl-d₃-5,5-dimethylhexane-1,1,1-d₃,

10a. Compound 10a was prepared from 29a according to

the procedure used (above) for 10. Residual protons for

the deuterated methyl groups could not be seen in the ¹H

NMR spectrum; it was assumed that deuterium incor-

poration was greater than 99%. A total of 0.64 g of 10a

1

were obtained (60%): H NMR (CDCl₃, 500 MHz), ꢀ

0.892 (s, 9H), 1.353 (m, 2H), 1.699 (m, 2H); ¹³C NMR

(CDCl₃, 125 MHz), ꢀ 29.30, 29.89, 31.43 (septet), 38.75,

40.75, 71.03.

2-Chloro-2,5,5-trimethylhexane-3,3-d₂, 10b. Com-

pound 10b was prepared using a modiﬁcation of the

procedure described by Shiner and Verbanic.⁷⁰Between

1 and 0.5ml of 29b was placed in a 1 dram (4ml) screw-

topped vial. A Pasteur pipette connected to a lecture

bottle of hydrogen chloride gas by means of Tygon

tubing was clamped into place above the vial so that the

tip of the pipette was almost touching the bottom of the

vial. Hydrogen chloride gas was gently bubbled into the

2,5,5-trimethyl-2-hexanol-3,3-d₂. After 5 min, the vial

became hot to the touch. After another 5 min, the liquid in

the vial developed a pinkish color and became cloudy.

After a total of 15 min, two layers were seen to form. The

bubbling was discontinued and the mixture was dissolved

2-Methyl-d₃-5,5-dimethyl1-2-hexanol-1,1,1-d₃, 29a.

Compound 29a was synthesized by the same procedure

as for 29, using acetone-d₆(99.9% D; Cambridge Isotope

Laboratories) instead of acetone. A 7.54 g amount of 29a

was obtained (60%), of which 1.43 g was puriﬁed by

preparative GC. Residual methyl protons could not be

J. Phys. Org. Chem. 12, 564–576 (1999)

g-SILYL-STABILIZED TERTIARY IONS

575

21. D. D. Davis and R. H. Black, J. Organomet. Chem. 82, C30–C34

(1974).

22. V. J. Shiner Jr and M. W. Ensinger, J. Am. Chem. Soc. 108, 842–

844 (1986).

23. V. J. Shiner Jr and M. W. Ensinger, J. Org. Chem. 55, 653–661

(1990).

24. D. J. Raber, W. C. Neal Jr, M. D. Dukes, J. M. Harris and D. L.

Mount, J. Am. Chem. Soc. 100, 8137–8146 (1978).

25. E. R. Davidson and V. J. Shiner Jr, J. Am. Chem. Soc. 108, 3135–

3137 (1986).

26. V. J. Shiner Jr, M. W. Ensinger and R. D. Rutkowske, J. Am. Chem.

Soc. 109, 804–809 (1987).

27. J. Coope and V. J. Shiner Jr, J. Org. Chem. 54, 4270–4271 (1989).

28. V. J. Shiner Jr, M. W. Ensinger and J. C. Huffman, J. Am. Chem.

Soc. 111, 7199–7205 (1989).

29. C. A. Grob and P. Sawlewicz, Tetrahedron Lett. 28, 951–952

(1987).

30. C. A. Grob, M. Gru¨ndel and P. Sawlewicz, Helv. Chim. Acta 71,

1502–1507 (1988).

in pentane, washed with water and aqueous sodium

hydrogen carbonate and again with water. The pentane

solution was then dried (K₂CO₃), and the solvent was

removed by rotary evaporation at ambient temperature.

The residue was then vacuum distilled on a molecular

still (lit.³²b.p. 46.5–47.5°C at 11mmHg) to give the

product. ¹H NMR analysis of residual methylene protons

showed deuterium incorporation to be approximately

90%; presumably some depletion occurred as a result of

1

addition of HCl to some elimination product. H NMR

(CDCl₃, 500 MHz), ꢀ 0.892 (s, 9H), 1.341 (s, 2H), 1.561

(s, 6H). ¹³C NMR (CDCl₃, 125 MHz), ꢀ 29.30, 29.89,

32.32, 38.62, 40.5 (multiplet ¹³C's for CD₂and CDH

compounds), 71.37. (Some small peaks believed to

correspond to the CDH compound were also seen in the

¹³C NMR spectrum at ꢀ 32.384, 38.706 and 71.5.)

31. C. A. Grob and A. Waldner, Helv. Chim. Acta 62, 1736–1744

(1979).

32. C. A. Grob and A. Waldner, Helv. Chim. Acta 62, 1854–1865

(1979).

33. E. Grunwald and S. Winstein, J. Am. Chem. Soc. 70, 846 (1948).

34. V. J. Shiner Jr, W. Dowd, R. D. Fisher, S. R. Hartshorn, M. A.

Kessick, L. Milakofsky and M. W. Rapp, J. Am. Chem. Soc. 91,

4838–4843 (1969).

35. R. D. Fisher, PhD Thesis, Indiana University, Bloomington, IN

(1971).

36. V. J. Shiner Jr, in Isotope Effects in Chemical Reactions. ACS

Monograph No. 167 edited by, C. J. Collins and N. S. Bowman,

Chapt. 2. Van Nostrand Reinhold, New York (1970).

37. X. Li and J. A. Stone, J. Am. Chem. Soc. 111, 5586 (1989).

38. D. J. Defrees, W. J. Hehre and D. E. Sunko, J. Am. Chem. Soc. 101,

2323–2327 (1979).

Acknowledgements

We thank Indiana University for support of this work. We

also thank Scott Ballentine and Mike Li for assistance in

the synthesis of 10 and deuterated analogs.

REFERENCES

39. C. K. Ingold, Structure and Mechanism in Organic Chemistry, 2nd

ed. Cornell University Press, New York (1969).

40. D. T. Stoelting and V. J. Shiner, J. Am. Chem. Soc. 115, 1695–1705

(1993).

41. B. L. Murr Jr, PhD Thesis, Indiana University, Bloomington, IN

(1961).

42. L. J. Tilley, PhD Thesis, Indiana University, Bloomington, IN

(1996).

43. K. J. Caserta, F. J. Holler, S. R. Crouch and C. G. Enke, Anal.

Chem. 50, 1534–1541 (1978).

44. M. W. Ensinger, PhD Thesis, Indiana University, Bloomington, IN

(1987).

45. R. Ensman, Indiana University Electronic Instrumentation Ser-

vices, personal communication.

46. R. Withnell, unpublished results.

47. F. P. Wilgis, unpublished work.

48. M. W. Rapp, PhD Thesis, Indiana University, Bloomington, IN,

(1968).

49. M. J. Tomasik, unpublished results.

50. C. R. Sporleder, Computer Systems Group, Indiana University,

Bloomington, IN, personal communication.

51. D. T. Stoelting unpublished results.

52. W. E. Buddenbaum, PhD Thesis, Indiana University, Blooming-

ton, IN (1964).

53. P. C. Vogel, PhD Thesis, Indiana University, Bloomington, IN

(1967).

54. A. R. Pinnick, PhD Thesis, Indiana University, Bloomington, IN

(1973).

55. K. D. Bowersox, unpublished results.

56. D. T. Stoelting, unpublished results.

1. (a) N. Shimizu, Rev. Heteroat. Chem. 9, 155–179 (1993); (b) J. B.

Lambert, Tetrahedron 46, 2677–2689 (1990).

2. F. K. Cartledge and J. P. Jones, Tetrahedron Lett. 2193–2196

(1971).

3. Y. Apeloig and A. Stanger, J. Am. Chem. Soc. 107, 2806–2807

(1985).

4. Y. Apeloig, R. Biton and A. Abu-Freih, J. Am. Chem. Soc. 115,

2522–2523 (1993).

5. N. Shimizu, E. Osajima and Y. Tsuno, Bull. Chem. Soc. Jpn. 64,

1145–1152 (1991).

6. M. A. Cook, C. Eaborn and D. R. M. Walton, J. Organomet. Chem.

29, 389–396 (1971).

7. M. Mishima, T. Ariga, Y., Tsuno, K. Ikenaga and K. Kikukawa,

Chem. Lett. 489–492 (1992).

8. S. G. Wierschke, J. Chandraseckhar and W. L. Jorgensen, J. Am.

Chem. Soc. 107, 1496–1500 (1985).

9. S. N. Ushakov and A. M. Itenberg, Zh. Obshch. Khim. 7, 2495

(1937).

10. J. B. Lambert, G.-T. Wang, R. B. Finzel and D. H. Teramura, J.

Am. Chem. Soc. 109, 7838–7845 (1987).

11. M. R. Ibrahim, and W. L. Jorgensen, J. Am. Chem. Soc. 111, 819–

824 (1989).

12. I. Fleming, Chemtracts Org. Chem. 6, 113–116 (1993).

13. J. B. Lambert, R. W. Emblidge and S. Malany, J. Am. Chem. Soc.

115, 1317–1320 (1993).

14. L. H. Sommer, E. Dorfman, G. M. Goldberg and F. C. Whitmore,

J. Am. Chem. Soc. 68, 488–489 (1946).

15. L. H. Sommer, R. E. Van Strien and F. C. Whitmore, J. Am. Chem.

Soc. 71, 3056–3061 (1949).

16. I. Fleming and J. P. Michael, J. Chem. Soc., Chem. Commun. 245–

247 (1978).

17. I. Fleming and J. P. Michael, J. Chem. Soc., Perkin Trans. 1 1549–

1556 (1981).

18. I. Fleming and S. K. Patel, Tetrahedron Lett. 22, 2321–2324

(1981).

57. M. W. Ensinger, unpublished results.

58. L. H. Sommer and N. S. Marans, J. Am. Chem. Soc. 72, 1935–1939

(1950).

59. F. C. Whitmore and L. H. Sommer, J. Am. Chem. Soc. 68, 481–484

(1946).

60. D. D. Perrin and W. L. F. Armarego, Puriﬁcation of Laboratory

Chemicals, 3rd ed. Pergamon Press, New York (1988).

61. S. C. Welch, J. M. Assercq, J. P. Loh and S. A. Glase, J. Org.

Chem. 52, 1440–1450 (1987).

19. I. Fleming, S. K. Patel and C. J. Urch, J. Chem. Soc., Perkin Trans.

1 115–124 (1989).

20. D. D. Davis and H. T. Johnson, J. Am. Chem. Soc. 96, 7576–7577

(1974).

J. Phys. Org. Chem. 12, 564–576 (1999)

576

L. J. TILLEY AND V. J. SHINER, JR

62. I. Fleming and J. Goldhill, J. Chem. Soc., Perkin Trans. 1 1493–

1498 (1980).

63. A. F. Thomas, Deuterium Labeling in Organic Chemistry.

Meredith, New York (1971).

64. A. Streitwieser Jr, L. Verbit and P. Stang, J. Org. Chem. 29, 3706

(1964).

65. D. Seyferth, R. E. Mammarella and H. A. Klein, J. Organomet.

Chem. 94, 1–8 (1980).

R. Wheeler, C. S. Miner Jr, L. H. Sutherland, R. B. Wagner, T. W.

Clapper, C. E. Lewis, A. R. Lux and A. H. Popkin, J. Am. Chem.

Soc. 63, 643–654 (1941).

68. L. Duhamel, P. Duhamel and A. Jarry, Bull. Soc. Chim. Fr. 1797–

1804 (1970).

69. A. Bra¨ndstro¨m, Acta Chem. Scand. 13, 613 (1959).

70. V. J. Shiner Jr and C. J. Verbanic, J. Am. Chem. Soc. 79, 369–373

(1957).

66. L. Schmerling, J. Am. Chem. Soc. 67, 1152–1154 (1945).

67. F. C. Whitmore, J. S. Whitaker, W. A. Mosher, O. N. Breivik, W.

J. Phys. Org. Chem. 12, 564–576 (1999)

Article Doi

γ-Silyl-stabilized tertiary ions? Solvolysis of 4-(trimethylsilyl)-2-chloro-2-methylbutane

DOI: 10.1002/(SICI)1099-1395(199907)12:7<564::AID-POC164>3.0.CO;2-2

Source and publish data:

Authors:

Article abstract of DOI:10.1002/(SICI)1099-1395(199907)12:7<564::AID-POC164>3.0.CO;2-2

Full text of DOI:10.1002/(SICI)1099-1395(199907)12:7<564::AID-POC164>3.0.CO;2-2

Products guided by the article

R&D Labs maybe for 245764-70-9

Relevant to this article

Hot Product