Mechanistic studies of intramolecular CH insertion reaction of arylnitrenes: Isotope effect, configurational purity and radical clock studies

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY

J. Phys. Org. Chem. 2005; 18: 9–20

Published online 28 June 2004 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/poc.837

Mechanistic studies of intramolecular CH insertion

reaction of arylnitrenes: isotope effect, configurational

purity and radical clock studies

Shigeru Murata,¹* Yasuhiro Tsubone,²Reina Kawai,²Daisuke Eguchi²and Hideo Tomioka²

¹Department of Basic Science, Graduate School of Arts and Sciences, The University of Tokyo, Meguro-ku, Tokyo 153-8902, Japan

²Chemistry Department for Materials, Faculty of Engineering, Mie University, Tsu, Mie 514-8507, Japan

Received 24 February 2004; revised 14 April 2004; accepted 28 April 2004

ABSTRACT: In order to reveal the mechanism of the intramolecular CH insertion of arylnitrenes, three experiments

were carried out: measurement of isotope effects, determination of the extent of conﬁgurational retention and radical

clock studies. Irradiation of the deuterium-substituted azide 4-d in an inert solvent exclusively afforded the indolines

5-h and 5-d, in which the kinetic isotope effect k_H/k_Don the intramolecular CH insertion of the nitrene was evaluated

as 12.6–14.7 at room temperature. A chiral chromatographic analysis of the indoline 11 obtained from the optically

active azide (S)-6 revealed that the enantiomeric purity of the starting azide was almost completely lost during the

intramolecular CH insertion of the photolytically generated nitrene (enantiomeric excess <10%). The thermolysis of

the azide 7 at 180 ^ꢀC mainly gave a mixture of the cyclopropyl ring-opened products 20–22, together with the

intramolecular CH insertion product with an intact cyclopropyl ring 19. On the basis of these observations, we

concluded that the intramolecular CH insertion of the nitrene proceeds primarily by the hydrogen abstraction–

recombination mechanism. We propose, however, a small contribution of the concerted mechanism to the

intramolecular CH insertion, based on the solvent dependence of the isotope effect and the extent of the

conﬁgurational retention. Copyright # 2004 John Wiley & Sons, Ltd.

KEYWORDS: nitrenes; hydrogen abstraction; deuterium isotope effect; optical purity; radical clock

INTRODUCTION

The reactivity of arylnitrenes has been of great interest

because of their practical applications, including photo-

resist systems and biochemical photoafﬁnity labeling.¹

The insertion of a photolytically generated nitrene into a

CH bond is believed to be an important bond-forming

reaction in many systems with practical applications,

while very few mechanistic studies of the CH insertion

of arylnitrenes have been reported. Although it is well

known that the CH insertion of arylnitrenes generated

photolytically in solutions is an unfavorable process,¹we

recently found that arylnitrenes can insert favorably into

a reactive CH bond, such as benzylic, which is located

close to the nitrenic center.⁵Thus, the photolysis of 2-(2-

phenylethyl)phenyl azide (4) in cyclohexane exclusively

afforded 2-phenylindoline (5), which was produced by

intramolecular insertion of the nitrene into a ꢀ-CH bond

of the 2-phenylethyl group.

Arylnitrenes are known to be short-lived intermediates

produced in photochemical and thermal decomposition

of aryl azides (for reviews on nitrene chemistry see, e.g.,

Ref. 1). The scheme of the photochemical decomposition

of phenyl azide (1) has been established, in which singlet

phenylnitrene (2S) is identiﬁed as the temperature-

dependent branching point (Scheme 1).²Recently, two

groups have independently reported the direct observa-

tion of 2S in ﬂuid solutions.³The activation energy for

the ring expansion of 2S to didehydroazepine (3) has

been estimated to be 6.2 kcal mol^ꢁ1(1 kcal ¼ 4.184 kJ)

and the absolute rate constant for intersystem crossing to

triplet nitrene (2T) has been directly determined to be

3.2 ꢂ 10⁶s^ꢁ1.⁴Moreover, Gritsan et al. established that

on the basis of the agreement between the experimental

and theoretical absorption spectra, 2S has an open-

shell electronic structure with two singly occupied non-

bonding orbitals.⁴

*Correspondence to: S. Murata, Department of Basic Science, Grad-

uate School of Arts and Sciences, The University of Tokyo, Meguro-

ku, Tokyo 153-8902, Japan.

In order to shed light on the mechanism of the

intramolecular CH insertion of arylnitrenes, we have

E-mail: cmura@mail.ecc.u-tokyo.ac.jp

Copyright # 2004 John Wiley & Sons, Ltd.

J. Phys. Org. Chem. 2005; 18: 9–20

10

S. MURATA ET AL.

contribute slightly but signiﬁcantly to the reaction.

Further, taking into account the electronic structure of

singlet phenylnitrene (2S) established recently,^4,9the spin

state of the nitrene involved in the CH insertion is

discussed.

RESULTS

Measurement of deuterium isotope effects

using the azide 4-d

Scheme 1

The synthetic route used to prepare the deuterium-

substituted azide 4-d is summarized in Scheme 2. Deu-

terium was introduced in the course of the reduction of 2-

nitrobenzyl phenyl ketone with LiAlD₄(Aldrich, 98%). It

designed three approaches. The ﬁrst is the measurement

of the kinetic isotope effect on the intramolecular CH

insertion using the deuterium-substituted azide 4-d (for a

preliminary report, see Ref. 6). If the CH insertion of the

nitrene generated from 4-d proceeds by the concerted

mechanism, small deuterium isotope effects are expected.

On the other hand, large values of k_H/k_D, which indicate a

high degree of CH bond cleavage in the transition state,

support the hydrogen abstraction–recombination me-

chanism. The second study is the determination of the

extent of conﬁgurational retention using optically active

(S)-2-(2-methylbutyl)phenyl azide [(S)-6].⁷A large ex-

tent of conﬁgurational retention during the insertion of

the nitrene into a CH bond at an asymmetric carbon atom

is expected in the concerted mechanism, while a loss of

optical purity of the intramolecular CH insertion product

gives a strong piece of evidence in support of the

hydrogen abstraction–recombination mechanism. Finally,

we have designed 2-(2-cyclopropylethyl)phenyl azide

(7), where the nitrene can be expected to react with a

ꢀ-CH bond of the 2-cyclopropylethyl group, to apply

the ‘cyclopropylcarbinyl clock’ approach (for recent

mechanistic studies with the cyclopropylcarbinyl clock

see, for example, Ref. 8). If the CH insertion proceeds by

the concerted mechanism, the cyclopropane ring is intact.

On the other hand, if the attack of the nitrene on the ꢀ-CH

bond proceeds by the hydrogen abstraction mechanism,

the formation of cyclopropyl ring-opened products is

expected because of the generation of the cyclopropyl-

carbinyl radical.

1

was conﬁrmed by the integration of H NMR that the

deuterium isotope purity of the starting reagent was held

in 2-(2-deuterio-2-phenylethyl)aniline (8-d), which was

converted to the azide 4-d in a usual manner.⁵

Irradiation of 4-d in cyclohexane with a high-pressure

mercury lamp through a Pyrex ﬁlter at 20 ^ꢀC gave a

mixture of the indolines 5-h and 5-d. The total yield of 5

was 52% based on the consumed starting azide. In the ¹H

NMR spectrum of the photoreaction mixture, a signal

assigned to one of the methylene protons of 5-h and 5-d

appeared at ꢁ 3.44 and a methine signal of 5-d appeared at

ꢁ 4.95. Thus, the ratio of 5-h to 5-d was readily obtained

1

by the integration of H NMR in the reaction mixture,

which was evaluated to be 14.7 ꢃ 0.3. The product ratio

remained constant during the photoreaction, although

continued irradiation caused a decrease in the total yield

of 5, which was due to the photochemical decomposition

of 5. No other products such as the corresponding aniline

8-d or azobenzene could be detected in the photoreaction

mixture.

It seems reasonable to assume that the kinetic isotope

effect k_H/k_Don the intramolecular insertion of the nitrene

generated photolytically from 4-d into a ꢀ-CH bond of

the 2-phenylethyl group is equal to the product ratio

[5-h]/[5-d]. Thus, we obtained an unusually large deuter-

ium isotope effect at room temperature. The values of the

isotope effects, and also the total yields of 5, gained in

the irradiation of 4-d under various conditions are sum-

marized in Table 1. As shown, the isotope effect obtained

In this paper, based on the results obtained from these

three approaches, we propose that the intramolecular CH

insertion of arylnitrenes generated by photolysis, in

addition to thermolysis, of the azides proceeds predomi-

nantly by the hydrogen abstraction–recombination me-

chanism, while the concerted mechanism could

Scheme 2

Copyright # 2004 John Wiley & Sons, Ltd.

J. Phys. Org. Chem. 2005; 18: 9–20

INTRAMOLECULAR CH INSERTION REACTION OF ARYLNITRENES

11

in the photolysis with the longer-wavelength light

( > 350 nm) was identical to that obtained with the

Pyrex-ﬁltered light (>300 nm) within the experimental

error. Further, it was found that the isotope effects were

dependent on the solvent employed in the photolysis:

slightly smaller values were obtained both in benzene and

in acetonitrile, compared with the value in cyclohexane.

In methanol, we could not obtain an accurate value of the

product ratio [5-h]/[5-d] owing to the minor formation of

5. Instead, irradiation of 4-d in methanol gave predomi-

nantly 2-(2-methoxy-2-phenylethyl)anilines, 9-h and 9-d

(69% in total), the formation of which has previously

reported in the photolysis of 4-h and rationalized in terms

of methanol trapping of the polarized biradical formed by

the intramolecular hydrogen abstraction of the nitrene.⁵

In an analogous manner, we obtained the ratio of 9-h to

9-d by integration of the ¹H NMR spectrum, from which

the deuterium isotope effect for the formation of 9 was

evaluated as 10.2 ꢃ 0.2, which is included in Table 1.

Extremely large deuterium isotope effects, k_H/

k_D¼ 12.6–14.7, observed for the formation of 5 suggest

that a quantum-mechanical tunneling mechanism contri-

butes to the reaction process.¹⁰It is known that a non-

linear plot of ln(k_H/k_D) versus T^ꢁ1is a general feature

in hydrogen transfer reactions which proceed by a quan-

tum-mechanical tunneling mechanism.^10–12Thus, the

temperature dependence of k_H/k_Dfor the formation of 5

was closely examined in cyclohexane (20–55 ^ꢀC) and in

acetonitrile (ꢁ14–59 ^ꢀC). Although we found that the

deuterium isotope effect k_H/k_Ddepended considerably on

the irradiation temperature, the isotope effect increasing

with decreasing irradiation temperature, Arrhenius plots

of the observed isotope effect, ln(k_H/k_D) versus T^ꢁ1, gave

a straight line within the experimental errors in the

limited temperature ranges employed in our experiments,

as illustrated in Fig. 1. To validate the contribution of a

Figure 1. Temperature dependence of the deuterium iso-

tope effects for the intramolecular CH insertion of the

nitrene generated by the photolysis of 4-d in cyclohexane

(*) and in acetonitrile (*)

tunneling mechanism, the measurement of the rate con-

stants for the intramolecular H and D abstraction of the

nitrene in a wide temperature range should be required,

which is difﬁcult under our experimental conditions,

unfortunately.

Moreover, we examined the deuterium isotope effect

on the intramolecular CH insertion of the thermally

generated nitrene. Thermolysis of 4-d in 1,2,4-trichlor-

obenzene at 180 ^ꢀC for 30 min afforded not only a

mixture of 5-h and 5-d (45% in total), but also 2-

phenylindole (10, 45%). The ratio of 5-h to 5-d was

evaluated as 5.9. However, we found that the ratio [5-h]/

[5-d], and also the product distribution, were very depen-

dent on the reaction time: prolonged thermolysis at

180 ^ꢀC for 1 h resulted in an increase in the yield of 10

to 55% with a decrease in that of 5, and an increase in the

ratio [5-h]/[5-d] to 7.8. The formation of the indole 10 is

reasonably interpreted in terms of the dehydrogenation of

5 in the course of the thermolysis. Further, the depen-

dence of the ratio [5-h]/[5-d] on the reaction time appears

to be attributable to the deuterium isotope effect on the

dehydrogenation process. Thus, because of the unex-

pected dehydrogenation reaction of 5, we failed to gain

an intrinsic value of the isotope effect on the CH insertion

of the nitrene generated by thermolysis of 4-d.

Table 1. Products and kinetic isotope effects obtained in the

photolysis of 4

Conditions

Yield (%)

a

Solvent

Light (nm)

5

9

k_H/k_D

Determination of the extent of configurational

retention using the azide (S)-6

Cyclohexane

Acetonitrile

Benzene

Methanol

>300

>350

>300

>350

>300

52

78

40

49

74

17

—

69

14.7 ꢃ 0.3

14.2 ꢃ 0.3

13.6 ꢃ 0.3

12.6 ꢃ 0.3

13.0 ꢃ 0.2

10.2 ꢃ 0.2^b

In 1964, Smolinsky and Feuer reported the solution-

phase thermolysis of optically active (S)-2-(2-methylbu-

tyl)phenyl azide [(S)-6], in which they concluded

that 65% of optical purity was retained during the

intramolecular CH insertion of the nitrene to give

^aAt 20–22 ^ꢀC.

^bObtained from 9.

Copyright # 2004 John Wiley & Sons, Ltd.

J. Phys. Org. Chem. 2005; 18: 9–20

12

S. MURATA ET AL.

2-ethyl-2-methylindoline (11).⁷However, their result was

based on the assumption that the speciﬁc rotation of 11

obtained from the vapor-phase thermolysis of (S)-6 is

nearly that of optically pure 11, which appears extremely

doubtful. We have found that the racemic indoline ( ꢃ )-

11 can be perfectly resolved by the use of high-perfor-

mance liquid chromatograph (HPLC) with a column

packed with optically active resin (DAISEL Chiralcel

OJ). By employing this method, we examined the extent

of the retention of enantiomeric purity during the intra-

molecular CH insertion of the nitrene generated by

photolysis, and also by thermolysis, of (S)-6.

The optically active azide 6 was synthesized from

commercially available optically active (S)-(ꢁ)-2-

methyl-1-butanol according to the synthetic route pre-

sented by Smolinsky and Feuer.⁷The absolute conﬁgura-

tion of 6 was reasonably assigned to be S, in agreement

with that of the starting material, because the asymmetric

carbon atom was intact in the course of synthesis. The

optical purity of 6 was evaluated as > 90% on the basis of

chiral chromatographic analysis of its precursor 2-(2-

methyl)butylaniline. Prior to conﬁgurational studies, the

structure and the distribution of photoproducts were

examined by using the racemic azide ( ꢃ )-6.⁷Irradiation

of ( ꢃ )-6 in cyclohexane with Pyrex-ﬁltered light gave

mainly 2-ethyl-2-methylindoline (11, 44%), which was

produced by the intramolecular CH insertion of the

nitrene at an asymmetric carbon atom. A minor photo-

product was also formed, which was tentatively identiﬁed

as 2-(2-methylenebutyl)aniline (12, 6%). Although the

unchanged starting material and tarry products were

readily removed by the use of gel permeation liquid

chromatography (GPLC) and thin-layer chromatography

(TLC), all attempts to isolate the indoline 11 from the

mixture of photoproducts were unsuccessful. However,

the analysis of 11 containing a small amount of 12 by the

use of HPLC with a chiral column gave two large, well-

separated peaks, together with small additional peaks, as

shown in Fig. 2. Since the intensities of the large two

peaks were identical within the experimental error

( ꢃ 2%), these peaks were reasonably assigned to the

enantiomers of the indoline 11. Thus, it was found that

contamination of the by-product 12 did not interfere with

the determination of enatiomeric purity of 11. The

photolysis of ( ꢃ )-6 was also carried out in methanol,

where a considerable amount of 2-(2-methoxy-2-methyl-

butyl)aniline (13, 20%) was produced together with 11

and 12 (43% and 3%, respectively). Furthermore, we

found that the photolysis of ( ꢃ )-6 in diethylamine

(DEA) afforded a small amount of 11 (3%), although

the 3H-azepine derivative 14, which is established as a

characteristic photoproduct of aryl azides in DEA,¹was

Figure 2. Chromatogram obtained by the injection of 11

produced by the photolysis of ( ꢃ )-6 in cyclohexane. Col-

umn, DAISEL Chiralcel OJ; eluent, hexane–2-propanol

(30:1); flow-rate, 0.6 ml min^ꢁ1; detection, 255 nm

predominantly obtained (24%). Dilution of DEA with

cyclohexane resulted in an increase in the yield of 11 with

a decrease in that of 14.

An enantiomeric excess of the indoline 11 produced in

the irradiation of the optically active azide (S)-6 was

directly determined by the chiral chromatographic ana-

lysis of 11 which was separated from the photoreaction

mixture by the use of GPLC and TLC. The values of the

enantiomeric excess, and also the yield of 11, obtained in

the irradiation of (S)-6 under various conditions are

summarized in Table 2. These values remained constant

during the photoreaction within the experimental error,

which excluded the possibility of racemization of 11

through the secondary photochemical cleavage of the

bond linking a nitrogen atom with an asymmetric carbon

atom. As shown in the table, the enantiomeric purity of

the starting azide 6 is found to be almost completely lost

during the intramolecular CH insertion of the nitrene

generated photolytically under all conditions studied. It

should be pointed out, however, that small but signiﬁcant

values of enantiomeric excess (ee) are observed, which

Copyright # 2004 John Wiley & Sons, Ltd.

J. Phys. Org. Chem. 2005; 18: 9–20

INTRAMOLECULAR CH INSERTION REACTION OF ARYLNITRENES

13

Table 2. Products and enantiomeric excess of 11 obtained

in the photolysis and thermolysis of (S)-6

Conditions

Solvent

Yield (%)

ee of 11 (%)

Light (nm)

11^a

Others

Cyclohexane

Methanol

>300

>250

>350

>300

50

—

46

3

11

3.2 ꢃ 1.0

10.9 ꢃ 0.3

2.6 ꢃ 1.0

6.2 ꢃ 1.4

9.4 ꢃ 0.6

6.5 ꢃ 0.2

b

20^c

DEA

24^d

Scheme 3

Cyclohexane–

DEA (9:1)

1,2,4-

6^d, 8^e

180 ^ꢀC

30

2^f

26.3 ꢃ 2.0

Trichlorobenzene

^aContaining a small amount of 12.

azide 7 and examined its photochemical and thermal

reactivities.

^bNot determined.

^c13.

^d14.

We could obtain the azide 7 according to the synthetic

route described in Scheme 3. Hydrogenation (PtO₂/

EtOH, 0 ^ꢀC) of ꢀ-cyclopropyl-2-nitrostyrene gave a mix-

ture of 2-(2-cyclopropylethyl)aniline (16) and 2-pentyla-

niline (6:1, 96% in total). After the conversion of the

aniline mixture into the corresponding azides, the azide 7

was successfully separated from the cyclopropyl ring-

opened azide by the use of GPLC.

^e2-(2-Methylbutyl)aniline.

^f15.

are clearly dependent on the solvent employed in the

photolysis. Although the absolute conﬁguration of the

predominantly produced enantiomer, which is eluted

later in the chromatographic analysis of 11, cannot be

determined, it is tentatively identiﬁed as R by assuming

that the intramolecular CH insertion proceeds partially

with a conﬁgurational retention of the starting azide. This

assumption was supported by the result obtained in the

thermolysis of (S)-6, which mainly afforded the enantio-

mer of 11 identical with that produced predominantly in

the photolysis. It has been reported that the thermolysis of

(S)-6 gives an intramolecular CH insertion product hav-

ing a retained conﬁguration.⁷

Thermolysis of the racemic azide ( ꢃ )-6 in 1,2,4-

trichlorobenzene at 180 ^ꢀC gave a mixture of 11 and 12

(21% and 9%, respectively). A trace amount of 2,3-

dimethyl-1,2,3,4-tetrahydroquinoline (15, 2%), the

formation of which was previously reported in the ther-

molysis of 6,⁷was isolated from the reaction mixture.

The enantiomeric excess of 11 obtained in the thermo-

lysis of (S)-6 was determined in an analogous manner,

which is included in Table 2.

Disappointingly, irradiation of 7 in cyclohexane with

Pyrex-ﬁltered light afforded a very complicated mixture.

Isolation and puriﬁcation of the photoproducts were

unsuccessful, although the azobenzene 17 was detected

1

in the crude reaction mixture, the H NMR spectrum of

which showed a characteristic doublet centered at ꢁ 7.63,

which was assigned to the proton at the ortho position of

—

N

N. This photoreactivity of 7 contrasts strikingly with

those of the azides 4 and 6, the photolysis of which in an

inert solvent exclusively gives the corresponding intra-

molecular CH insertion product of the nitrene. These

results are possibly interpreted in terms of a greater

binding energy of the ꢀ-CH bond of 7, compared with

that of the benzylic and the tertiary ꢀ-CH bond of 4 and 6,

respectively. Curiously, although no CH insertion pro-

ducts were obtained in the irradiation of 7 in methanol,

the methoxylated aniline 18 was isolated in 23% yield,

together with the aniline 16 (4%). This observation

indicates that the nitrene generated from 7 can attack a

ꢀ-CH bond of the 2-cyclopropylethyl group with the aid

of a polar solvent. The participation of the solvent would

lead to a polarized transition state for the hydrogen

abstraction, which directly gives the intermediate having

a large zwitterionic character. Hence it is reasonable to

think that the biradical having a cyclopropylcarbinyl

moiety cannot participate in the formation of the meth-

oxylated aniline 18.

Photochemical and thermal decomposition

of the azide 7

The ring opening of cyclopropylcarbinyl radical to the 3-

butenyl radical has been used as a mechanistic probe for

reactions thought to involve free radicals.⁸The formation

of ring-opened products from the material having a

cyclopropylcarbinyl group gives unambiguous evidence

for the formation of a free radical at the position adjacent

to the cyclopropane ring. In order to apply this approach

to gain information about the mechanism of the intramo-

lecular CH insertion of arylnitrenes, we designed the

Copyright # 2004 John Wiley & Sons, Ltd.

J. Phys. Org. Chem. 2005; 18: 9–20

14

S. MURATA ET AL.

Thus, it is found that the photolysis of the azide 7 is not

It should be pointed out that three of the products, the

pyrroloindoline 20, the tetrahydroquinoline 21 and the

dimeric pyrroloquinoline 22, are deﬁnitely cyclopropyl

ring-opened products, which can be produced through the

cyclopropylcarbinyl radical rearrangement. Hence this

observation provides unambiguous evidence supporting

that the free radical having a cyclopropylcarbinyl radical

moiety is generated during the thermolysis of 7. Further-

more, it seems reasonable to expect that the intramole-

cular attack on a ꢀ-CH bond of the nitrene produced

thermally from 7 is responsible for the generation of the

cyclopropylcarbinyl radical.

suitable for radical clock studies. It seems reasonable to

assume that an elevated temperature is favorable for the

intramolecular insertion to the CH bond having a greater

binding energy. Therefore, the thermolysis of 7 was

examined next. A solution of 7 in 1,2,4-trichlorobenzene

was heated at 180 ^ꢀC. After separation by the use of

GPLC and TLC, four reaction products were obtained.

The ﬁrst product was identiﬁed as 2-cyclopropylindoline

1

(19, 9%), the H NMR spectrum of which showed the

cyclopropyl ring protons in a highly shielded area (ꢁ 0.2–

1.1). The second product was assigned to the pyrroloin-

doline 20 (7%). The ¹H NMR spectrum of 20 showed the

absence of a cyclopropyl ring and a hydrogen attached to

a nitrogen atom. The ¹H–¹H COSY spectrum of 20

exhibited a correlation of the methine proton at ꢁ 3.91–

3.96 with both the benzylic protons (ꢁ 2.96 and 3.15–

3.22) and the methylene protons located remote from a

DISCUSSION

Mechanism of intramolecular CH insertion

of arylnitrenes

nitrogen atom (ꢁ 1.29–1.38 and 1.81–1.93). The structure

of 20 was conﬁrmed by the agreement of its H and ¹³

1

C

Photochemical decomposition of azides. The

photolysis of the azides 4 and 6 in an inert solvent

exclusively afforded the intramolecular CH insertion

product 5 and 11, respectively. Extremely large deuter-

ium isotope effects, k_H/k_D¼ 12.6–14.7, are obtained in

the photolysis of 4-d at room temperature. These results

present an unambiguous evidence supporting that the

intramolecular CH insertion proceeds not by the con-

certed mechanism, but by the hydrogen abstraction–

recombination mechanism. The same conclusion can be

drawn from the results obtained in the conﬁgurational

studies of the photochemistry of (S)-6, where the extent

of the conﬁgurational retention during the intramolecular

insertion of the nitrene into a CH bond at an asymmetric

carbon atom is found to be <10%. Thus, based on the

results obtained in these two independent experiments, it

is safely concluded that the intramolecular insertion of

the photolytically generated nitrene into the reactive CH

bond located close to the nitrenic center proceeds

primarily by the hydrogen abstraction–recombination

mechanism.

It should be pointed out that the deuterium isotope

effect and the extent of the conﬁgurational retention are

dependent slightly but signiﬁcantly on the solvent em-

ployed in the photolysis, as shown in Tables 1 and 2. The

largest value of the kinetic isotope effect, and also the

smallest value of the enantiomeric retention, are observed

in cyclohexane, while a smaller value of the isotope effect

and a larger value of the enantiomeric retention compared

with those in cyclohexane are observed in an electron-

donating solvent. We propose that the solvent effects

observed are attributed to a small contribution of the

concerted mechanism to the intramolecular CH insertion

of the nitrene. It could be assumed that in the concerted

mechanism, the insertion of the nitrene into the CH bond

is initiated by the charge-transfer interaction of the

occupied ꢂ orbital of the CH bond with the vacant orbital

on the nitrogen atom. On the other hand, the nitrene

NMR data with those of 20 reported by Ziegler and

Jeroncic.¹³The third product, which showed a set of

signals due to a vinyl group in its ¹H NMR spectrum, was

identiﬁed as 3-vinyltetrahydroquinoline (21, 6%). The

position of the vinyl group was determined from its

¹H–¹H COSY spectrum, in which the sequence of pro-

—

tons —CH —CH(CH CH )—CH — was estab-

—

2

1

lished. The fourth product, the H, H–¹H COSY and

¹³C NMR spectra of which exhibited its unsymmetirical

dimeric structure having an intact 2-(cyclopropylethyl)-

phenyl group, was assigned to the pyrroloquinoline 22

(8%). Prolonged heating of the reaction mixture simply

resulted in a decrease in the yields of these reaction

products, which excluded the possibility of thermal

interconversion among the reaction products. Judging

1

from the H NMR spectrum of the reaction mixture

obtained after the thermolysis, there were no other

products formed in an appreciable yield. The use of a

hydrogen-donating solvent, such as decahydronaphtha-

lene, in the thermolysis caused great difﬁculty in separat-

ing the reaction products from polymeric products

derived from the solvent. We ﬁnally conﬁrmed by sepa-

rate experiments that the cyclopropyl ring was unchanged

under conditions employed in the thermolysis of 7.

Copyright # 2004 John Wiley & Sons, Ltd.

J. Phys. Org. Chem. 2005; 18: 9–20

INTRAMOLECULAR CH INSERTION REACTION OF ARYLNITRENES

15

responsible for the intramolecular hydrogen abstraction

should have an open-shell electronic structure with two

singly occupied molecular orbitals on the nitrogen atom.

Thus, the interaction of the vacant orbital on the nitrogen

atom with a lone pair of the solvent employed in

the photolysis would stabilize the electronic structure of

the nitrene participating in the former mechanism.

Alternatively, a more polar transition state of the con-

certed CH insertion, compared with that of the hydrogen

abstraction, would be stabilized in an electron-donating

solvent. It seems reasonable to think that these effects

increase the contribution of the concerted mechanism in

an electron-donating solvent. Because of a low degree of

CH bond cleavage in the transition state of the concerted

CH insertion, an increase in the contribution of the

concerted mechanism results in a decrease in the kinetic

isotope effect and an increase in the extent of the

conﬁgurational retention. The spin state and electronic

structure of the reactive intermediate responsible for the

concerted mechanism are extensively discussed in a

following section.

Scheme 4

estimated for the ring-opening reaction leads us to assume

that the concerted CH insertion of the thermally generated

nitrene gives 19 (Scheme 4). Again, this assumption is in

harmony with the conclusion that the concerted mechan-

ism contributes relatively largely to the intramolecular CH

insertion of the thermally generated nitrene.

Thermal decomposition of azides. Previously,

Smolinsky and Feuer presented the ﬁrst report on the

partial retention of optical activity during the intramole-

cular CH insertion of the nitrene generated thermally from

(S)-6.⁷Now, we have strictly determined the extent of the

conﬁgurational retention as 26.3% in 1,2,4-trichloroben-

zene at 180^ꢀC. Thus, we could conclude that the intra-

molecular CH insertion reaction of the thermally generated

nitrene also proceeds by the hydrogen abstraction–recom-

bination mechanism predominantly. It should be pointed

out, however, that the value of the conﬁgurational retention

is considerably larger than those obtained in the photolysis,

which would be reasonably explained in terms of a larger

contribution of the concerted mechanism.

This interpretation is not inconsistent with the results

obtained in the thermolysis of 7, where the cyclopropyl

ring-opened products 20–22 are predominantly produced.

This observation deﬁnitely indicates the generation of the

cyclopropylcarbinyl radical 23 during the thermolysis, the

formation of which is reasonably explained in terms of

the intramolecular hydrogen abstraction of the thermally

generated nitrene (Scheme 4). Moreover, it should be

noted that we have obtained a signiﬁcant amount of the

intramolecular CH insertion product with an intact

cyclopropyl ring 19. Arrhenius parameters for the cyclo-

propylcarbinyl radical rearrangement of 1-cyclopropy-

lethyl radical have been estimated to be log(A/

Although the detailed process of the formation of the

cyclopropyl ring-opened products 20–22 has not been

elucidated yet, we tentatively propose a scheme involving

the intramolecular cyclization of the biradical 24 resulting

from the cyclopropylcarbinyl opening of 23 (Scheme 4).

The intramolecular attack of the aminyl radical on the 2-

position of the 2-pentenylene moiety in 24 would give the

biradical with a ﬁve-membered ring, which would be a

precursor of the pyrroindoline 20. If the intramolecular

cyclization occurs at the 3-position, the biradical with a

six-membered ring would be formed, which could give not

only 21 by the recombination, followed by the re-cleavage

of the resulting four-membered ring, but also 22 by the

reaction with the unchanged azide 7. It seems reasonable

to think that a variety of the interconversion between

reactive intermediates is allowed at the reaction tempera-

ture of 180 ^ꢀC, and that the isolation of the products 20–22

is primarily attributable to their high thermostability.

Although isotope labeling studies, and also theoretical

studies of the relative stability of intermediate biradicals,

are required to validate this scheme, it should be empha-

sized again that these products 20–22 obtained in the

thermolysis of the azide 7 are deﬁnitely produced through

the cyclopropylcarbinyl radical rearrangement.

s

^ꢁ1) ¼ 13.15 and E_a¼ 7.5 kcal mol^{ꢁ1 14}

.

Assuming that

these parameters are applicable in a high temperature

range, the rate constant for the ring opening of the 1-

cyclopropylethyl radical is evaluated to be ca 3 ꢂ 10⁹s^ꢁ1

at 180^ꢀC. Although the possibility that 19 is produced

through the simple recombination of the biradical 23

generated by the intramolecular hydrogen abstraction

cannot be ruled out yet, an extremely large rate constant

Spin state and electronic structure of the nitrene

responsible for the intramolecular CH insertion

The CH insertion of photolytically generated arylcar-

benes has been fully investigated, and it is established

that the reaction proceeds generally by the concerted

Copyright # 2004 John Wiley & Sons, Ltd.

J. Phys. Org. Chem. 2005; 18: 9–20

16

S. MURATA ET AL.

mechanism involving a singlet state carbene (for reviews

on carbene chemistry see, for example, Ref. 15). Recently,

based on low stereoselectivities and large deuterium

isotope effects (k_H/k_D¼ 4–8), Kirmse and co-workers

reported that a triplet-state carbene participated in the

intramolecular CH insertion with steric constraints in its

transition state.¹⁶The assumption underlying the discus-

sion on the spin state of arylcarbenes involved in the

reaction is a difference in the electronic structure between

singlet- and triplet-state carbenes. It is generally accepted

that singlet carbene has a closed-shell electronic structure

with one ﬁlled and one vacant non-bonding orbital,

whereas triplet carbene has an open-shell electronic

structure with two singly occupied non-bonding orbitals.

In contrast to arylcarbenes, recent theoretical and experi-

mental studies have clearly demonstrated that singlet, in

addition to triplet, phenylnitrene (2) has an open-shell

electronic structure.^4,9Assuming that the alkyl substituent

at the ortho-position has no inﬂuence on the electronic

structure of 2, the intramolecular CH insertion of the

nitrene is expected to proceed by the hydrogen abstrac-

tion–recombination mechanism, regardless of its spin

state. Our observations described in the preceding sections

are primarily consistent with this expectation. It should be

emphasized, however, that we have demonstrated a con-

tribution of the concerted mechanism, which cannot be

expected from an open-shell electronic structure, to the

intramolecular CH insertion of the nitrene generated by

photolysis, in addition to thermolysis. Here we have a

discussion on the spin state and electronic structure of the

reactive intermediate involved in the CH insertion which

proceeds by the concerted mechanism.

We propose the following four possible mechanisms

for the concerted intramolecular CH insertion of nitrenes.

The ﬁrst is a contribution of the electronically excited

singlet state. The ﬁrst electronically excited singlet state

of phenylnitrene (2S) is expected to have a closed-shell

electronic structure, which is designated as the ¹A₁

state.^4,9It is unlikely, however, that the singlet nitrene

with the ¹A₁state is thermally populated, even under the

conditions employed in the thermolysis, because the

energy gap between the ground singlet state (¹A₂) and

¹A₁is estimated to be ca 18 kcal mol^ꢁ1by both CASSCF

calculation⁹and the modern CASPT2 method.⁴The

second possibility is the participation of an electron

transfer from the CH bond to the electronegative nitrogen

atom in the intramolecular CH insertion. The interaction

of the electron localized on the CH bond with one of the

singly occupied orbitals of the nitrenic center would lead

to the polarized three-centered transition state of the CH

insertion, in which the extent of the CH bond cleavage is

relatively low. The polarized transition state of the

intramolecular CH insertion of the nitrene has been

demonstrated by the formation of the methoxylated

anilines 9 and 13 in the photolysis of the azides 4 and

6, respectively, in methanol. Moreover, the participation

of an electron transfer in the nitrene chemistry has been

proposed recently.^17,18The third possibility is an inter-

action between the azide moiety and the CH bond in a

ground state of the azide. The electron-donating inter-

action of the occupied molecular orbital localized on

the CH bond with the unoccupied molecular orbital of the

azide moiety could lead to the transition state of the

concerted intramolecular CH insertion without passing

through the nitrene. The interaction between the azide

moiety and an electron-donating group substituted at

the ortho-position of phenyl azide has been established

in the thermolysis of various aryl azides.^1,19Finally, we

propose that the electronically excited azides could be a

candidate for the intermediate responsible for the con-

certed intramolecular CH insertion of the photochemi-

cally generated nitrene. (In the photochemistry of

diazirines, the participation of their electronically excited

state in the product formation has been proposed.²⁰) It

seems that a single electronically excited state of the

azide was involved in the photolysis with the Pyrex-

ﬁltered light ( > 300 nm), because the isotope effect and

the enantiomeric excess obtained in the photolysis of 4-d

and (S)-6, respectively, with the light of > 350 nm are

identical with those obtained with the Pyrex-ﬁltered light.

As shown in Table 2, however, the use of the shorter-

wavelength light ( > 250 nm) afforded a signiﬁcantly

large value of the conﬁgurational retention during the

intramolecular CH insertion of the nitrene generated from

(S)-6. This observation appears to be interpretable in

terms of the contribution of the higher excited state of

(S)-6 to the intramolecular CH insertion.

At the present stage, we cannot specify the mechanism

of the concerted intramolecular CH insertion of the

nitrenes, which is observed as a minor process in our

experiments. Further experimental and theoretical studies

are required to elucidate the details of the reactivity of

arylnitrenes.

CONCLUSION

In order to gain experimental information about the

mechanism of the intramolecular CH insertion of arylni-

trenes, we carried out three experiments: measurement of

deuterium isotope effects, determination of conﬁgura-

tional purities and cyclopropylcarbinyl radical clock

studies. On the basis of the results obtained, we con-

cluded that the intramolecular CH insertion of the ni-

trenes proceeds primarily by the hydrogen abstraction–

recombination mechanism. The spin state of the nitrenes

involved in the hydrogen abstraction cannot be deter-

mined, because recent theoretical and experimental

Copyright # 2004 John Wiley & Sons, Ltd.

J. Phys. Org. Chem. 2005; 18: 9–20

INTRAMOLECULAR CH INSERTION REACTION OF ARYLNITRENES

17

studies have established that singlet in addition to triplet,

phenylnitrene (2) has an open-shell electronic

structure.^4,9We should point out, however, a small but

signiﬁcant contribution of the concerted mechanism to the

intramolecular CH insertion, which cannot be explained

by an open-shell electronic structure of the nitrenes.

1-ethanone in 20ml of dry diethyl ether was added 174 mg

(4.15 mmol) of LiAlD₄(Aldrich, 98%) at 0 ^ꢀC. The

reaction mixture was stirred for 1 h at room temperature.

After the successive addition of MeOH and 10% sulfuric

acid to the mixture, and the organic material was extracted

with diethyl ether. The extract was washed with water and

dried over Na₂SO₄. The solvent was removed under

reduced pressure to afford 972 mg (96%) of 1-deuterio-

1-phenyl-2-(2-nitrophenyl)ethanol. A solution of the re-

sulting alcohol (815 mg, 3.34 mmol) and 50 mg of 4-

toluenesulfonic acid monohydrate in benzene (50 ml)

was reﬂuxed for 1 h. The mixture was washed successively

with saturated NaHCO₃solution and water and dried over

Na₂SO₄. The solvent was removed under reduced pres-

sure, and the residue was developed on a silica gel column

with hexane–CH₂Cl₂(3:1) to give 680 mg (90%) of 1-

deuterio-1-phenyl-2-(2-nitrophenyl)ethene. The identity

EXPERIMENTAL

¹H NMR spectra were recorded at 270 or 500 MHz. ¹³C

NMR spectra were recorded at 126 MHz. GC analyses

were performed on a column prepared from 5% silicone

OV-17 on Diasolid L (5.0 mm ꢂ 1.0 m). HPLC analysis

was performed on a Hitachi L-6000 system. Gel permea-

tion liquid chromatography (GPLC) was carried out on a

JASCO HLC-01 high-performance liquid chromatograph

equipped with a Shodex GPC H-2001 column. Thin-layer

chromatography (TLC) was carried out on Merck Kie-

1

and purity of the material were established by H NMR

spectrum: yellow granules; m.p. 66–67 ^ꢀC; ¹H NMR

(CDCl₃), ꢁ 7.29–7.49 (5H, m), 7.53–7.63 (3H, m), 7.77

(1H, d, J ¼ 7.9 Hz), 7.97 (1H, d, J ¼ 8.3 Hz).

selgel 60 PF₂₅₄

.

Materials. The synthesis of ( ꢃ )- and (S)-2-(2-methyl-

butyl)phenyl azide (6) has been reported previously.⁷

2-(2-Deuterio-2-phenylethyl)phenyl azide (4-d). A solu-

tion of 650 mg (2.88 mmol) of 1-deuterio-1-phenyl-2-(2-

nitrophenyl)ethene in 50 ml of acetic acid was stirred

with 50 mg of 10% Pd/C under a hydrogen atmosphere

overnight at room temperature. The solid was ﬁltered and

washed several times with acetic acid. The solvent was

removed under reduced pressure. To the residue were

added 20 ml of water and the mixture was neutralized

with Na₂CO₃. The organic material was extracted with

diethyl ether and the extract was washed with water and

dried over Na₂SO₄. The solvent was removed under

reduced pressure to give 562 mg (99%) of 2-(2-deuterio-

2-phenylethyl)aniline (8-d). The deuterium isotope purity

Preparation of 2-(2-deuterio-2-phenylethyl)phenyl azide

(4-d). 1-Phenyl-2-(2-Nitrophenyl)-1-ethanone. To a solu-

tion of 3.68 ml of benzaldehyde (36.1 mmol) and 4.29 ml

of 2-nitrotoluene (36.4 mmol) in 50 ml of freshly distilled

DMSO was added a solution of sodium ethoxide in EtOH

(0.43 ^M, 10 ml). The mixture was stirred for 2 days at

room temperature and concentrated to ca 5 ml by dis-

tillation under reduced pressure. After the addition of

20 ml of water, the reaction mixture was extracted with

diethyl ether. The organic layer was washed with water

and dried over Na₂SO₄. The solvent was removed under

reduced pressure to afford 3.88 g (44%) of 1-phenyl-2-(2-

nitrophenyl)ethanol. The resulting alcohol (3.60 g,

14.8 mmol) was dissolved in 50 ml of acetone. To the

solution was added dropwise an acidic chromium re-

agent, which was prepared by the dilution of aqueous

solution of chromium(VI) oxide (5.5 ^M, 3.6 ml) with

sulfuric acid (4.4 ^M, 7.4 ml), until the starting alcohol

was no longer detected by TLC. The mixture was further

stirred for 10 min and concentrated to ca 10 ml by

evaporation. After the addition of 30 ml of water, the

organic material was extracted with CH₂Cl₂. The extract

was washed with water and dried over Na₂SO₄. The

solvent was removed under reduced pressure and the

residue was developed on a silica gel column with

hexane–CH₂Cl₂(1:1) to give 2.12 g (59%) of 1-phenyl-

2-(2-nitrophenyl)-1-ethanone. The identity and purity of

the material were established by ¹H NMR spectrum: light

1

of 8-d was found by NMR integration to be > 95%; H

NMR (CDCl₃), ꢁ 2.78 (2H, d, J ¼ 8.3 Hz), 2.92 (1H, d,

J ¼ 8.3 Hz), 6.66–6.77 (2H, m), 7.02–7.07 (2H, m), 7.19–

7.33 (5H, m). To a solution of 562 mg (2.83 mmol) of 8-d

in 15 ml of dioxane were added 15 ml of 6 ^Nsulfuric acid.

The mixture was cooled to 0–5 ^ꢀC and a solution of

195 mg (2.83 mmol) of NaNO₂in 2 ml of water was

added dropwise to the mixture. The reaction mixture

was stirred for 30 min at this temperature. The mixture

was added dropwise to a solution of 3.68 g of NaN₃in

22 ml of water with stirring at room temperature. After

the addition, the reaction mixture was stirred for 2 h. The

organic material was extracted with CH₂Cl₂and the

extract was washed with water and dried over Na₂SO₄.

The solvent was removed under reduced pressure and the

residue was developed on a silica gel column with hexane

to give 431 mg (68%) of the azide 4-d. The identity and

1

yellow granules; m.p. 64–65 ^ꢀC; H NMR (CDCl₃), ꢁ

4.74 (2H, s), 7.38 (1H, d, J ¼ 6.3 Hz), 7.43–7.67 (5H, m),

8.05 (2H, d, J ¼ 6.9 Hz), 8.17 (1H, d, J ¼ 8.3 Hz); EIMS,

m/z 165 (14), 105 (100).

1

purity of the azide were established by H NMR and IR

spectra, which were completely identical with those of 4-

h,⁵except that the ¹H NMR spectrum showed only three

methylene protons at ꢁ 2.85; EIMS, m/z 224 (M^þ, 3), 196

(89), 195 (100), 119 (20).

1-Deuterio-1-phenyl-2-(2-nitrophenyl)ethene. To a solu-

tion of 1.00 g (4.15mmol) of 1-phenyl-2-(2-nitrophenyl)-

J. Phys. Org. Chem. 2005; 18: 9–20

18

S. MURATA ET AL.

Preparation of 2-(2-cyclopropylethyl)phenyl azide (7). (2-

nitrobenzyl)triphenylphosphonium bromide. A solution

of 15.0 g of 2-nitrotoluene (109 mmol), 17.6 g of N-

bromosuccinimide (99 mmol) and 0.14 g of benzoyl

peroxide in 66 ml of CCl₄was reﬂuxed for 5 h. The

precipitate was ﬁltered and washed with hot CCl₄. The

solvent was removed under reduced pressure. The resi-

dual 2-nitrobenzyl bromide was dissolved in 40 ml of

benzene and to the solution were added 26.0 g (99 mol) of

triphenylphosphine. The reaction mixture was allowed to

stand overnight at room temperature. A viscous material

was produced that solidiﬁed slowly. The solid was

ﬁltered, washed with diethyl ether and dried in vacuo to

afford 36.0 g (76%) of (2-nitrobenzyl)triphenylphospho-

nium bromide. This phosphonium salt was insoluble in

organic solvents and used without further puriﬁcation:

yellow granules; m.p. 226–228 ^ꢀC; IR (KBr disk), 1540,

(2H, brs), 6.66–6.75 (2H, m) and 6.99–7.06 (2H, m) in

the ¹H NMR spectrum. Although ¹H NMR and GC

analyses revealed that the product contained small

amounts of 2-pentylaniline (ca 15%), the product was

used without further puriﬁcation. To a solution of 418 mg

of the crude aniline in 16 ml of dioxane were added

16.8 ml of 6 ^Nsulfuric acid. The mixture was cooled to

0–5 ^ꢀC and a solution of 188 mg (2.72 mmol) of NaNO₂

in 3.5 ml of water was added dropwise to the mixture.

The reaction mixture was stirred for 1 h at this tempera-

ture. The mixture was added dropwise to a solution of

338 mg of NaN₃in 7.4 ml of water with stirring at room

temperature. After the addition, the reaction mixture was

stirred for 2 h. The organic material was extracted with

CH₂Cl₂and the extract was washed with water and dried

over Na₂SO₄. The solvent was removed under reduced

pressure and the residue was developed on a silica gel

column with hexane to give 332 mg (68%) of the crude

azide. The azide 7 was separated from 2-pentylphenyl

azide and puriﬁed by the use of GPLC. The identity and

1440, 1340, 1120, 1000 cm^ꢁ1

.

ꢀ-Cyclopropyl-2-nitrostyrene. Under an N₂atmosphere,

to a suspension of 5.36 g of (2-nitrobenzyl)triphenylpho-

sphonium bromide (11.2 mmol) in 30 ml of dry benzene

was added a solution of n-butyllithium in hexane (1.69 ^M,

6.64 ml, 11.2 mmol) at 0 ^ꢀC with vigorous stirring. After

stirring for 1 h at this temperature, to the reaction mixture

was added slowly a solution of 561 mg of cyclopropane-

carbaldehyde (8.00 mmol) in 8 ml of dry benzene. The

mixture was stirred overnight at room temperature. Water

was added to the mixture and the organic material was

extracted with diethyl ether. The extracted was washed

with water and dried over Na₂SO₄. The solvent was

removed under reduced pressure and the residue was

developed on a silica gel column with hexane–CH₂Cl₂

(5:1) to give 609 mg (39%) of ꢀ-cyclopropyl-2-nitro-

styrene as a mixture of cis and trans isomers (cis:

trans ¼ 2.5). The identity and purity of the material

were established from the ¹H NMR spectrum: light

1

purity of the azide were established by H NMR and IR

1

spectra: light yellow liquid; H NMR (CDCl₃), ꢁ 0.00–

0.09 (2H, m), 0.38–0.45 (2H, m), 0.64–0.77 (1H, m), 1.45

(2H, q, J ¼ 7.5 Hz), 2.65 (1H, t, J ¼ 7.8 Hz), 7.03–7.23

(4H, m); IR (NaCl), 2110, 1280, 750 cm^ꢁ1

.

Irradiation for preparative experiments. A solu-

tion of 20–30 mg of an azide in 20–30ml of a solvent was

placed in a Pyrex tube, purged with N₂for 10 min and

irradiated with a 300 W high-pressure mercury lamp for

60 min at room temperature. The solvent was removed

under reduced pressure and the residue was separated by

GPLC with CHCl₃eluent or preparative TLC. The identity

and purity of photoproducts were established from the ¹H

and ¹³C NMR spectra. The yield of products described in

the text was determined by isolation on the basis of the

reacted material. 2-Phenylindoline (5), 2-(2-phenylethy-

l)aniline (8) and 2-(2-methoxy-2-phenylethyl)aniline (9),

which were obtained by the irradiation of the azide 4, were

identiﬁed by comparison of the spectroscopic data with

those of authentic material.⁵

1

yellow liquid; H NMR (CDCl₃), ꢁ 0.40–0.44 (2H, m),

0.68–0.75 (2H, m), 1.41–1.50 (1H, m), 5.11 (1H, dd,

J ¼ 11.2, 10.6 Hz), 6.54 (1H, d, J ¼ 11.2 Hz), 7.18–7.58

(3H, m), 7.91 (1H, d, J ¼ 9.2 Hz) for the cis isomer; ꢁ

0.45–0.51 (2H, m), 0.77–0.84 (2H, m), 1.51–1.63 (1H,

m), 5.66 (1H, dd, J ¼ 15.5, 9.2 Hz), 6.85 (1H, d,

J ¼ 15.5 Hz), 7.18–7.58 (3H, m), 7.77 (1H, d,

J ¼ 8.9 Hz) for the trans isomer; EIMS, m/z 189 (M^þ,

6), 172 (13), 144 (32), 128 (47), 115 (100).

Irradiation of ( ꢃ )-6. Irradiation of ( ꢃ )-6 in cyclohex-

ane, followed by separation by preparative TLC with

hexane–CH₂Cl₂(1:1), gave 2-ethyl-2-methylindoline

(11), which was identiﬁed by comparison of the data

with those of authentic material.⁷The indoline 11 was

contaminated by small amounts of the product, which

showed signals at ꢁ 1.06 (3H, t, J ¼ 7.3 Hz), 2.04 (2H, q,

J ¼ 7.3 Hz), 3.30 (2H, s), 4.73 (1H, s) and 4.87 (1H, s) in

2-(2-Cyclopropylethyl)phenyl azide (7). A solution of ꢀ-

cyclopropyl-2-nitrostyrene (267 mg, 1.41 mmol) in 10 ml

of absolute ethanol was stirred in the presence of 10 mg

of PtO₂under a hydrogen atmosphere for 2.5 h at 0 ^ꢀC.

The solid was ﬁltered and washed several times with

ethanol. The solvent was removed under reduced pres-

sure to give 218 mg of the product containing 2-(2-

cyclopropylethyl)aniline (16), which showed signals at

ꢁ 0.05–0.10 (2H, m), 0.42–0.49 (2H, m), 0.70–0.83 (1H,

m), 1.51 (2H, q, J ¼ 7.5 Hz), 2.60 (2H, t, J ¼ 7.8 Hz), 3.60

1

the H NMR spectrum. The minor product was tenta-

tively identiﬁed as 2-(2-methylenebutyl)aniline (12). At-

tempts to isolate 11 from the mixture were unsuccessful.

Irradiation of ( ꢃ )-6 in methanol gave 2-(2-methoxy-2-

1

methylbutyl)aniline (13), together with 11. 13: oil; H

NMR (CDCl₃), ꢁ 0.98 (3H, t, J ¼ 7.5 Hz), 1.09 (3H, s),

1.52–1.59 (1H, m), 1.66–1.74 (1H, m), 2.47 (1H, d,

J. Phys. Org. Chem. 2005; 18: 9–20

INTRAMOLECULAR CH INSERTION REACTION OF ARYLNITRENES

19

J ¼ 14.2 Hz), 2.99 (1H, d, J ¼ 14.2 Hz), 3.15 (3H, s),

6.63–6.70 (2H, m), 6.69 (1H, d, J ¼ 7.3 Hz). 7.04 (1H,

t, J ¼ 7.6 Hz); ¹³C NMR (CDCl₃), ꢁ 8.2, 21.6, 30.1, 41.8,

48.8, 79.4, 116.1, 117.9, 124.3, 127.2, 132.6, 146.8. In

order to avoid the oxidation of photoproducts by mole-

cular oxygen,^5,21the photoreaction mixture obtained by

the irradiation of ( ꢃ )-6 in diethylamine was worked up

as follows: the photolyzed solution was added dropwise

to a ﬂask containing 50 ml of boiling MeOH under an N₂

atmosphere. The mixture was reﬂuxed for 1 h. After

cooling, the solvent was removed and the residue was

separated by GPLC to afford 2-(diethylamino)-3-(2-

methylbutyl)-3H-azepine (14), together with 11. 14: oil;

¹H NMR (CDCl₃), ꢁ 0.77–0.89 (6H, m), 1.12–1.16 (9H,

m), 1.21–1.29 (1H, m), 1.31–1.36 (1H, m), 3.38–3.46

(4H, m), 4.10–4.15 (1H, m), 5.08–5.15 (1H, m), 5.58–

5.61 (1H, m), 6.23–6.27 (1H, m), 7.04 (1H, d,

J ¼ 7.6 Hz); ¹³C NMR (CDCl₃), ꢁ 11.2, 11.3, 13.3,

13.4, 19.0, 19.9, 29.2, 29.6, 29.9, 30.1, 32.5, 32.7, 39.7,

39.8, 43.9, 108.2, 108.3, 115.8, 116.3, 128.0, 128.2,

139.3, 146.8, 147.3.

evaporation of the solvent, the residue was separated by

GPLC with CHCl₃as eluent. The identity and purity of

1

products were established from the H NMR spectrum.

The yield of products described in the text was deter-

mined by isolation. 2-Phenylindole (10)⁵obtained by the

thermolysis of 4, and 2,3-dimethyl-1,2,3,4-tetrahydroqui-

noline (15)⁷obtained from ( ꢃ )-6, were identiﬁed by

comparison of the spectroscopic data with those of

authentic material. Thermolysis of 7, followed by

GPLC separation, gave four products, the structures of

1

which were determined from the H, ¹³C NMR and

¹H–¹H COSY spectra. 2-Cyclopropylindoline (19): oil;

¹H NMR (CDCl₃), ꢁ 0.22–0.29, (2H, m), 0.48–0.54 (2H,

m), 1.03–1.10 (1H, m), 2.87–2.95 (1H, m), 3.11–3.19

(2H, m), 3.92 (1H, brs), 6.62 (1H, d, J ¼ 7.6 Hz), 6.69

(1H, t, J ¼ 7.3 Hz), 7.01 (1H, t, J ¼ 7.6 Hz), 7.09 (1H, d,

J ¼ 7.3 Hz); ¹³C NMR (CDCl₃), ꢁ 2.2, 3.1, 16.7, 36.0,

65.0, 109.1, 118.5, 124.7, 127.2, 128.7, 150.7; EIMS, m/z

159 (M^þ, 100), 130 (47), 118 (57). 2,3,3a,4-Tetrahydro-

1H-pyrrolo[1,2-a]indole (20):¹³oil; ¹H NMR (CDCl₃), ꢁ

1.29–1.38 (1H, m), 1.81–1.93 (3H, m), 2.96 (1H, dd,

J ¼ 16.2, 2.4 Hz), 3.15–3.22 (2H, m), 3.42–3.46 (1H, m),

3.91–3.96 (1H, m), 6.60 (1H, d, J ¼ 7.9 Hz), 6.76 (1H, t,

J ¼ 7.3 Hz), 7.08–7.12 (2H, m); ¹³C NMR (CDCl₃), ꢁ

25.8, 31.3, 33.9, 52.3, 65.3, 111.0, 119.3, 124.9, 127.6,

129.9, 154.7. 3-Vinyl-1,2,3,4-tetrahydroquinoline (21):

Irradiation of 7. Irradiation of 7 in methanol, followed by

GPLC separation, gave the aniline 16 and 2-(2-cyclopro-

pyl-2-methoxyethyl)aniline (18). 18: oil; ¹H NMR

(CDCl₃), ꢁ 0.05–0.10 (1H, m), 0.41–0.48 (1H, m),

0.49–0.52 (1H, m), 0.63–0.69 (1H, m), 0.77–0.84 (1H,

m), 2.72–2.76 (1H, m), 2.85 (2H, d, J ¼ 4.9 Hz), 3.38

(3H, s), 6.67–6.72 (2H, m), 7.03 (2H, m); ¹³C NMR

(CDCl₃), ꢁ 5.0, 13.8, 37.7, 56.6, 87.2, 115.6, 117.9,

124.5, 126.8, 130.9, 145.5.

1

oil; H NMR (CDCl₃), ꢁ 2.60–2.67 (1H, m), 2.68 (1H,

t, J ¼ 10.1 Hz), 2.82–2.85 (1H, m), 3.07 (1H, t,

J ¼ 9.8 Hz), 3.34–3.37 (1H, m), 3.88 (1H, brs), 5.07

(1H, d, J ¼ 10.4 Hz), 5.15 (1H, d, J ¼ 17.1 Hz), 5.87

(1H, ddd, J ¼ 17.1, 10.4, 6.4 Hz), 6.50 (1H, d,

J ¼ 7.6 Hz), 6.62 (1H, t, J ¼ 7.3 Hz), 6.96–6.99 (2H, m);

¹³C NMR (CDCl₃), ꢁ 32.9, 36.2, 46.9, 113.9, 114.6,

117.1, 120.5, 126.9, 129.6, 140.2, 144.1. 2-[2-(2-Cyclo-

propylethyl)phenyl]-1H-2,3,3a,4,9,9a-hexahydropyrrolo

[2,3-b]quinoline (22): oil; ¹H NMR (CDCl₃), ꢁ 0.09–0.12

(2H, m), 0.47–0.51 (2H, m), 0.76–0.82 (1H, m), 1.49–

1.55 (2H, m), 1.65–1.72 (1H, m), 2.35–2.41 (1H, m), 2.60

(2H, t, J ¼ 7.8 Hz), 3.14 (1H, dd, J ¼ 16.2, 2.5 Hz), 3.25

(1H, dd, J ¼ 16.5, 9.2 Hz), 3.32–3.37 (1H, m), 3.56–3.60

(1H, m), 3.61–3.66 (1H, m), 3.65 (1H, brs), 3.76–3.80

(1H, m), 6.60 (1H, d, J ¼ 8.2 Hz), 6.66 (1H, d,

J ¼ 7.9 Hz), 6.70 (1H, t, J ¼ 7.3 Hz), 6.80 (1H, t,

J ¼ 7.3 Hz), 7.06–7.11 (3H, m), 7.14 (1H, t, J ¼ 7.6 Hz);

¹³C NMR (CDCl₃), ꢁ 4.7, 11.0, 31.1, 33.1, 33.4, 34.0,

51.5, 58.1, 70.8, 110.8, 111.1, 117.4, 119.8, 125.0, 126.5,

127.0, 127.7, 129.1, 129.1, 145.1, 154.4.

Irradiation for analytical experiments. In a typical

run, a solution of an azide (3 mg) in a solvent (3 ml) was

placed in a Pyrex tube, purged with N₂for 10 min and

irradiated with a 300 W high-pressure mercury lamp. The

consumption of the material (<50%) and the yield of the

photoproducts were determined by the integration of

1

the H NMR spectrum in the crude reaction mixture.

Identiﬁcation of the product was established by the

agreement of the ¹H NMR spectra with those of authentic

samples. The ratio of 5-h to 5-d obtained by the irradia-

tion of 4-d was also determined by the integration of the

¹H NMR spectrum. The optical purity of 11 obtained by

the irradiation of (S)-6 was determined by HPLC with a

column packed with DAISEL Chiralcel OJ using hexane–

2-propanol (30:1) as eluent.

REFERENCES

Thermolysis. Under an N₂atmosphere, 1 ml of 1,2,4-

trichlorobenzene in a ﬂask was heated at 180 ^ꢀC. A

solution of an azide (10 mg) in 1,2,4-trichlorobenzene

(1 ml) was added dropwise to the ﬂask with stirring. The

mixture was stirred for 30 min at this temperature. After

cooling to room temperature, the mixture was developed

on a silica gel column with hexane to remove the solvent.

Elution with CH₂Cl₂gave a mixture of products. After

1. (a) Iddon B, Meth-Cohn O, Scriven EFV, Suschitzky H, Gallagher

PT. Angew. Chem., Int. Ed. Engl. 1979; 18: 900–917; (b) Scriven

EFV. In Reactive Intermediates, Abramovitch RA (ed.). Plenum

Press: New York, 1982; (c) Scriven EFV (ed.). Azides and

Nitrenes. Academic Press: New York, 1984; (d) Wentrup C.

Reactive Molecules. Wiley: New York, 1984; Chapt. 4.

2. Leyva E, Platz MS, Persy G, Wirz J. J. Am. Chem. Soc. 1986; 108:

3783–3790.

J. Phys. Org. Chem. 2005; 18: 9–20

20

S. MURATA ET AL.

3. (a) Born R, Burda C, Senn P, Wirz J. J. Am. Chem. Soc. 1997; 119:

5061–5062; (b) Gritsan NP, Yuzawa T, Platz MS. J. Am. Chem.

Soc. 1997; 119: 5059–5060.

4. Gritsan NP, Zhu Z, Hadad CM, Platz MS. J. Am. Chem. Soc. 1999;

121: 1202–1207.

5. Murata S, Yoshidome R, Satoh Y, Kato N, Tomioka H. J. Org.

Chem. 1995; 60: 1428–1434.

6. Murata S, Tsubone Y, Tomioka H. Chem. Lett. 1998; 549–550.

7. Smolinsky G, Feuer BI. J. Am. Chem. Soc. 1964; 86: 3085–

3088.

8. (a) Caldwell RA, Zhou L. J. Am. Chem. Soc. 1994; 116: 2271–

2275; (b) Beckwith ALJ, Bowry VW. J. Am. Chem. Soc. 1994;

116: 2710–2716; (c) Paul GC, Gajewski JJ. Tetrahedron Lett.

1995; 36: 8549–8552.

9. (a) Hrovat DA, Waali EE, Borden WT. J. Am. Chem. Soc. 1992;

114: 8698–8699; (b) Karney WL, Borden WT. J. Am. Chem. Soc.

1997; 119: 1378–1387.

Ruzicka J, Leyva E, Platz MS. J. Am. Chem. Soc. 1992; 114:

897–905.

12. Caldin EF. Chem. Rev. 1969; 69: 135–156.

13. Ziegler FE, Jeroncic LO. J. Org. Chem. 1991; 56: 3479–3486.

14. Bowry VW, Lusztyk J, Ingold KU. J. Am. Chem. Soc. 1991; 113:

5687–5698.

15. (a) Kirmse W (ed.). Carbene Chemistry. Academic Press: New

York, 1971; (b) Moss RA, Jones M Jr (eds). Carbenes, vols 1 and

2. Wiley: New York, 1973, 1975.

¨

16. (a) Kirmse W, Ozkir IS. J. Am. Chem. Soc. 1992; 114: 7590–7591;

¨

(b) Kirmse W, Ozkir IS, Schnitzler D. J. Am. Chem. Soc. 1993;

115: 792–793; (c) Kirmse W, Schnitzler D. Tetrahedron Lett.

1994; 35: 1699–1702.

17. (a) Liang T-Y, Schuster GB. J. Am. Chem. Soc. 1986; 108: 546–

547; (b) Liang T-Y, Schuster GB. J. Am. Chem. Soc. 1987; 109:

7803–7810; (c) Murata S, Mori Y, Satoh Y, Yoshidome R,

Tomioka H. Chem. Lett. 1999; 597–598.

10. (a) Brunton G, Griller D, Barclay LRC, Ingold KU. J. Am. Chem.

Soc. 1976; 98: 6803–6811; (b) Malatesta V, Ingold KU. J. Am.

Chem. Soc. 1981; 103: 3094–3098; (c) Engel PS, Chae W-K,

Baughman SA, Marschke GE, Lewis ES, Timberlake JW,

Luedtke AE. J. Am. Chem. Soc. 1983; 105: 5030–5034.

11. (a) Senthilnathan VP, Platz MS. J. Am. Chem. Soc. 1980; 102:

7637–7643; (b) Platz MS, Senthilnathan VP, Wright BB,

McCurdy CW Jr. J. Am. Chem. Soc. 1982; 104: 6494–6501; (c)

18. Murata S, Nakatsuji R, Tomioka H. J. Chem. Soc., Perkin Trans. 2

1995; 793–799.

19. Dyall LK, Kemp JE. J. Chem. Soc. B 1968; 976–979.

20. (a) White WR III, Platz MS. J. Org. Chem. 1992; 57: 2841–2846;

(b) Modarelli DA, Morgan S, Platz MS. J. Am. Chem. Soc. 1992;

114: 7034–7041.

21. Sundberg RJ, Suter SR, Brenner M. J. Am. Chem. Soc. 1972; 94:

513–520.

J. Phys. Org. Chem. 2005; 18: 9–20

Article Doi

Mechanistic studies of intramolecular CH insertion reaction of arylnitrenes: Isotope effect, configurational purity and radical clock studies

DOI: 10.1002/poc.837

Source and publish data:

Authors:

Article abstract of DOI:10.1002/poc.837

Full text of DOI:10.1002/poc.837

Products guided by the article

R&D Labs maybe for 53334-33-1

Relevant to this article

Hot Product