1-[(E)-2-arylethenyl]-2,2-diphenylcyclopropanes: Kinetics and mechanism of rearrangement to cyclopentenes

Helvetica Chimica Acta – Vol. 94 (2011)

1359

1-[(E)-2-Arylethenyl]-2,2-diphenylcyclopropanes: Kinetics and Mechanism

of Rearrangement to Cyclopentenes

by Johann Mulzer^a), Rolf Huisgen*^b), Vladimir Arion^c), and Reiner Sustmann^d)

^a) Institut fꢀr Organische Chemie der Universitꢁt Wien, Wꢁhringer Strasse 38, A-1090 Wien

^b) Department Chemie der Ludwig-Maximilians-Universitꢁt, Butenandt-Strasse 5 – 13,

D-81377 Mꢀnchen

^c) Institut fꢀr Anorganische Chemie der Universitꢁt Wien, Wꢁhringer Strasse 42, A-1090 Wien

^d) Institut fꢀr Organische Chemie der Universitꢁt Duisburg-Essen, Universitꢁtsstrasse 5, D-45141 Essen

´

Dedicated to Paul von Rague Schleyer, friend and mentor, on the occasion of his 80th birthday

Kinetic measurements for the thermal rearrangement of 2,2-diphenyl-1-[(E)-styryl]cyclopropane

(22a) to 3,4,4-triphenylcyclopent-1-ene (23a) in decalin furnished DH⁼_isom¼ 31.0 ꢀ 1.2 kcal mol^ꢁ1and

DS⁼_isom¼ ꢁ6.0 ꢀ 2.6 e.u. The lowering of DH⁼by 20 kcal mol^ꢁ1, compared with the rearrangement of the

vinylcyclopropane parent, is ascribed to the stabilization of a transition structure (TS) with allylic

diradical character. The racemization of (þ)-(S)-22a proceeds with DH⁼_rac¼ 28.2 ꢀ 0.8 kcal mol^ꢁ1

and DS⁼_rac¼ ꢁ5 ꢀ 2 e.u., and is at 1508 106 times faster than the rearrangement. Seven further

1-(2-arylethenyl)-2,2-diphenylcyclopropanes 22, (E)- and (Z)-isomers, were synthesized and charac-

terized. The (E)-compounds showed only modest substituent influence in their k_rac(at 119.48) and k_isom

(at 159.38) values. The lack of solvent dependence of rate opposes charge separation in the TS, but a

linear relation of log k_racwith log p.r.f., i.e., partial rate factors of radical phenylations of ArH, agrees with

a diradical TS. The ring-opening of the preponderant s-trans-conformation of 22 gives rise to the 1-exo-

phenylallyl radical 26 that bears the diphenylethyl radical in 3-exo-position, and is responsible for

racemization. The 1-exo-3-endo-substituted allylic diradical 27 arises from the minor s-gauche-

conformation of 22 and is capable of closing the three- or the five-membered ring, 22 or 23, respectively.

The discussion centers on the question whether the allylic diradical is an intermediate or merely a TS.

Quantum-chemical calculations by Houk et al. (1997) for the parent vinylcyclopropane reveal the lack of

an intermediate. Can the conjugation of the allylic diradical with three Ph groups carve the well of an

intermediate?

1. Introduction. – 1.1. Facts and Interpretations. Rearrangements of the C-backbone

of organic compounds have aroused the chemistꢂs interest early on. The thermal

rearrangement of vinylcyclopropane to cyclopentene, observed by two groups in 1960

[1][2], was a late-comer in a noble old family. For one of the discoverers, E. Vogel, it

was worth only a footnote in a review. Soon, the formal simplicity of this conversion

initiated a multitude of studies on scope and mechanism.

The first assumption of a diradical intermediate with allylic stabilization found

solid ground in kinetics and thermochemistry. Flowers and Frey independently

observed the rearrangement and, in 1961, published a gas kinetic study, which

furnished an activation energy E_A¼ 49.6 kcal mol^ꢁ1[3]. Decades later, this value was

rectified to 51.7 ꢀ 0.5 kcal mol^ꢁ1[4]. The thermal breakup of the cyclopropane ring

ꢃ 2011 Verlag Helvetica Chimica Acta AG, Zꢀrich

1360

Helvetica Chimica Acta – Vol. 94 (2011)

requires E_A¼ 63.0 kcal mol^ꢁ1, as measured via the cis > trans equilibration of

[1,2-²H₂]cyclopropane and reported in a classic paper by Schlag and Rabinovitch [5].

The difference of ca. 12 kcal mol^ꢁ1could well be accounted for by the allylic

stabilization of an intermediate 3 (Scheme 1); 14.0 – 14.5 kcal mol^ꢁ1for the conjugation

energy of the parent allyl radical resulted from a recent determination [6].

Scheme 1

Two conformations of 2, s-trans > s-gauche, produce different diradicals on ring-

opening. They bear the side chain on the exo-position (in 1) or on the endo-side (in 3) of

the allylic system; only 3 is expected to yield cyclopentene (4) by radical recombina-

tion.

With the orbital-symmetry rules for concerted processes, Woodward and Hoffmann

opened a new epoch of organic chemistry in 1965 (for a review, see [7]). In their

systematics of sigmatropic reactions, the rearrangement of vinylcyclopropane offered

one of the rare examples of 1,3 alkyl migration. The obvious course, suprafacial with

retention (sr), is ꢄforbiddenꢂ, as is likewise antarafacial with inversion (ai), but two other

pathways, antarafacial with retention (ar) and suprafacial with inversion (si), would

obey the orbital symmetry rules, all this under the proviso of concertedness. The

transition structures (TSs) of these processes look awkward to a varying degree.

Woodward and Hoffmann mentioned the vinylcyclopropane rearrangement only in

passing; they did not refer to the question whether or not the rearrangement would in

reality be concerted. However, Woodwardꢂs authority may have played a role, when

subsequently the orbital-symmetry-controlled, concerted courses dominated the

discussion.

Willcott and Cargle observed in 1967 that cis-vinyl[2-²H₁]cyclopropane equilibrates

with the trans-isomer at 3608; the loss of stereospecificity was at least five times as fast

as the irreversible rearrangement to cyclopentene [8]. Doering and Sachdev (1974/75)

studied optically active cis-2-isopropenylcyclopropane-1-carbonitrile and the trans-

Helvetica Chimica Acta – Vol. 94 (2011)

1361

isomer. At 2188, the rearrangement competed with racemization and diastereomeriza-

tion; the cyclopentene retained some enantiomeric excess (ee) making it possible to

determine the whole set of rate constants [9]. The ꢄcontinuous diradicalꢂ enriched the

vocabulary of interpretation.

Baldwin and co-workers gained valuable insight by investigating optically active

trans-2,2’-disubstituted vinylcyclopropanes (for a review, see [10]). The 2,2’-dimethyl

compound 5 at 2968 furnished all four cyclopentenes, i.e., two pairs of enantiomers [11]

(Scheme 2). The kinetic data were expressed in partial rate constants of four concerted

processes, in which the ꢄWH-allowedꢂ ones, si and ar (73%), dominated over the ꢄWH-

forbiddenꢂ sr and ai (27%). Compared with the high stereospecificities observed for

electrocyclic reactions, the preference is meager. The inclination toward orbital-

symmetry control is even smaller for [2,2’-²H₂] labeling: si þ ar reach only 53% [12].

Scheme 2

Tests with other 2,2’-substituents likewise showed a bias toward si þ ar [10]. In a

brilliant investigation, Asuncion and Baldwin in 1995 dealt with the rearrangement of

optically active trans-2-phenyl-1-[(E)-styryl]cyclopropane and the cis-isomer; the rates

of the competing enantiomerization and diastereomerization were likewise determined

[13]. The deconvolution of the complex kinetics was highly demanding and disclosed a

preference for the enantiomers of trans-3,4-diphenylcyclopent-1-ene over those of the

cis-isomer. Thus, not the benefit of orbital symmetry, but sober thermodynamics appear

to govern the steric course of the rearrangement. ꢄIt seems rather more plausible to view

these vinylcyclopropane to cyclopentene rearrangements as passing through alternative

kinetically competitive diradical transition structures.ꢂ [13].

1.2. Computational Studies. Concertedness of rearrangement, however, reentered

through another door: computational chemistry. The experimental endeavors were

accompanied by calculations, notably by Houk et al. (1992; [14]), who were interested

in 1,3-sigmatropic shifts. In 1997, Houk et al. [15], as well as Davidson and Gajewski [16],

reported on calculations of the vinylcyclopropane rearrangement by UB3LYP/6-31G*,

confirmed by CAS-SCF(4,4) studies. A later extended version by Houk and co-workers

included calculations of 2’-(tert-butyl)- and several methylated vinylcyclopropanes

[17].

An intrinsic reaction coordinate, starting from the s-gauche conformation of 2 –

without a TS for ring-opening – reached a slanting high plane at ca. 44 kcal mol^ꢁ1

ascending to the TS of 1,5-cyclization at 46.9 kcal mol^ꢁ1. Structures on this plane return

to s-gauche-2 without a barrier. The mentioned TS is ꢄessentially pure diradical in

characterꢂ [17] and corresponds to the endo-allylic 3 in Scheme 1. Yet, the reaction is

concerted, and the TS is located on the si pathway. In the TS, the two C,C bonds of the

1362

Helvetica Chimica Acta – Vol. 94 (2011)

allylic radical have similar bond lengths; the distance C(1)ꢁC(3) is 2.489 ꢅ and that of

C(1)ꢁC(5) 2.681 ꢅ. The question arises whether these long bonds furnish enough

energy for keeping the system on its concerted course. The energy of concert cannot be

high.

In the minimum-energy conformation of the endo-allylic diradical 3, the allyl

resonance keeps C(1) to C(4) in a common plane with only C(5) being mobile.

Incipient bonding C(1) ··· C(3) requires less deformation than needed for C(1) ··· C(5).

The molecular model shows that a certain strain must be overcome to achieve the

conformation of Houkꢂs TS with its nearly equal distances for C(1)ꢁC(3) and

C(1)ꢁC(5); this TS should be at a higher energy level than the conformationally relaxed

diradical 3. Why does the latter not occur as an intermediate?

´

In Houkꢂs splendid expose [17], a closely related TS can lead to enantiomerized

vinylcyclopropanes. That is achieved with less activation energy on another reaction

coordinate, which converts s-trans-2 into the exo-substituted allylic radicals; a C_s-

symmetric TS (like 1 in Scheme 1), located at 43.5 kcal mol^ꢁ1, enters – with or without

rotation – into the 1,3-cyclization. Should 1 not be an intermediate?

1.3. New Contributions. As described in a preceding paper, 1-(2-arylethenyl)-2,2-

diphenylcyclopropanes became easily accessible [18]. It was shown that three Ph

groups indeed stabilize the TS and bring down the temperature of the vinylcyclopro-

pane rearrangement. It is not a far-fetched idea that electronic and steric effects of

three Ph groups should cooperate in generating a deeper trough for the diradical

intermediate and make it interceptible. It seemed rewarding to study the kinetics of

racemization and rearrangement of oligophenylated vinylcyclopropanes and to

examine the influence of substitutents and solvents on the rate. The pertinent

computational study by Sustmann et al. reported in the following paper [19] is not free

of surprises.

2. Results. – 2.1. Preparation of 1-[2-Arylethenyl]-2,2-diphenylcyclopropanes¹).

Recently, we described the preparation of vinylcyclopropane derivatives 8 by the

interaction of diazo(diphenyl)methane with 1-substituted buta-1,3-dienes [18]

(Scheme 3). The 1,3-cycloaddition to the unsubstituted C¼C bond of 6 at room

Scheme 3

1

)

The IUPAC-conform systematic names of all compounds, mentioned in this work, are given in

Exper. Part in parentheses.

Helvetica Chimica Acta – Vol. 94 (2011)

1363

temperature is followed by a fast elimination of N₂from the initially formed 4,5-

dihydro-3H-pyrazole 7; the (E)-configuration at the disubstituted C¼C bond of 6 is

retained. Although convenient and productive, the method leads only to racemic

material.

Free of this disadvantage is a pathway via 2,2-diphenylcyclopropane-1-carboxylic

acid (11), the optical resolution of which has been reported [20]. The reduction to

aldehyde 14 and subsequent Wittig olefination should provide model compounds 22a –

22h for the study of racemization and rearrangement. Synthesis and reactions were first

tested with racemic material.

The ester 10 is easily available from methyl acrylate and diazo(diphenyl)methane

[20]. For the reduction of 11 to the aldehyde 14, the acid chloride 12 was converted to

the aziridide 13, which was treated with LiAlH₄, a known procedure [21]. We obtained

49% for the sequence 11 ! 14. The cycloaddition of diazo(diphenyl)methane to

acrolein gave 70% of rac-14.

The phosphonium salts for the Wittig reaction were prepared from benzyl-type

halides and Ph₃P. Even the somewhat exotic 4-[(E)-styryl]benzyl bromide (15) [22] was

reacted with Ph₃P in refluxing xylene to give 16. Anthracene-9-aldehyde (17) was

subjected to a modified LeuckartꢁWallach reaction to give 9-[(dimethylamino)methyl]-

anthracene (18). The quaternary ammonium salt 19, obtained with MeI, was converted

to the phosphonium iodide 20 with Ph₃P in refluxing BuOH (72% for 17 ! 20).

Various base systems were used for the deprotonation of the phosphonium salts and

the in situ reactions of the (arylmethylidene)(triphenyl)phosphoranes 21 with the

carbaldehyde 14 (Scheme 4). Our slight preference for EtONa in EtOH (Variant B; cf.

Exper. Part) was based on a somewhat higher purity of the products. The problem – and

failure – of a general stereocontrol of the Wittig reaction is as old as this valuable olefin

synthesis. The 1-(2-arylethenyl)-2,2-diphenylcyclopropanes 22 were formed as mixtures

of (E)- and (Z)-isomers of 22a – 22h in total yields of 69 – 89%, and the (Z)-content

varied from 56 to 93% (Table 1). In organic synthesis, (E)-olefins are usually in higher

demand than the (Z)-isomers. However, in our study the substantial amounts of (Z)-22

were a fringe benefit, since their thermal reactions are noteworthy [23].

The rather laborious separation of (E)- and (Z)-22 was performed by thick-layer

chromatography (TLC) on 2-mm silica gel on glass plates with light petroleum as

1364

Helvetica Chimica Acta – Vol. 94 (2011)

Scheme 4

Table 1. Properties of rac-1-(2-Arylethenyl)-2,2-diphenylcyclopropanes

Formula No. Ar

Yield M.p.

[%] [8]

d(HꢁC(1’)) d(HꢁC(2’)) ³J(1’,2’) C¼C (str.)

[ppm]

[Hz]

[cm^ꢁ1

]

(E)-22a

(E)-22b

(E)-22c

(E)-22d

(E)-22e

(E)-22f

(E)-22g

(E)-22h

(Z)-22a

(Z)-22b

(Z)-22c

(Z)-22d

(Z)-22e

(Z)-22f

(Z)-22g

(Z)-22h

Ph

31

22

30

28

18

60 – 61

88 – 89

77 – 79

5.46

5.47

5.49

5.34

5.68

5.48

5.17

5.47

4.94

4.90

4.99

4.85

5.15

4.96

5.16

5.47

6.55

6.58

6.50

6.52

6.62

6.68

15.8

16.0

15.5

16.0

11.5

11.6

11.5

11.4

11.0

1638

1643

1639

1640

1633

1639

1642

1635

1640

1627

no signal

1629

1628

1635

no signal

4-MeꢁC₆H₄

3-ClꢁC₆H₄

4-MeOꢁC₆H₄

4-NO₂ꢁC₆H₄

61 – 63

98 – 100

144 – 145

108 – 110

135 – 136

88 – 89

62 – 63

67 – 68

83 – 84

112 – 113

125 – 127

107 – 108

oil

4-[(E)-Styryl]ꢁC₆H₄32

a

Naphthalen-1-yl

Anthracen-9-yl

Ph

20

22

39

61

45

41

59

)

a

6.39

6.37

6.33

6.41

6.39

6.86

4-MeꢁC₆H₄

3-ClꢁC₆H₄

4-MeOꢁC₆H₄

4-NO₂ꢁC₆H₄

4-[(E)-Styryl]ꢁC₆H₄57

Naphthalen-1-yl

Anthracen-9-yl

62

58

a

)

^a) Signal among those of aromatic H-atoms.

mobile phase and UV check on zone separation²). The (Z)-structures always showed

higher R_fvalues than the (E) forms.

The vinylcyclopropanes 22 were crystalline compounds (with one exception), and

the (E,Z)-assignment was infallibly based on the coupling constant of the vinylic H-

atoms: J(1’,2’) ¼ 15.5 – 16.0 for (E)-22a – 22h and 11.0 to 11.6 Hz for (Z)-22a – 22h. The

chemical shifts, d(HꢁC(1’)) and d(HꢁC(2’)), are higher for (E)- than for (Z)-isomers

(Table 1). The IR HC¼CH stretching frequency at ca. 1640 cm^ꢁ1is regularly a bit

higher for (E)- than for (Z)-22, but changes of intensity are considerable.

2

)

The experimenter regarded his work as an orgy in TLC.

Helvetica Chimica Acta – Vol. 94 (2011)

1365

Furthermore, the out-of-plane (oop) deformation frequencies for (E)- and (Z)-

HC¼CH occur in different regions [24]. However, aromatic oop frequencies populate

these regions also, sharply reducing the value of this criterion here.

2.2. Optically Active 1-[2-(E)-Arylethenyl]-2,2-diphenylcyclopropanes and Their

(Z)-Isomers. Walborsky and Hornyak [20] achieved the optical resolution of the

carboxylic acid 11 via the brucine salt in moderate yield. With the crystallization of the

quinine salt of (þ)-11 from acetone, the same laboratory improved the procedure [25].

On treatment with HCl, the quinine salt produced optically pure (þ)-11 with [a]_D²⁵

¼

þ226 (CHCl₃). The quinine salt of (ꢁ)-11 from the mother liquor was impure, but,

on treatment with brucine in acetone, could be converted to diastereomerically pure

brucine salt and with HCl to (ꢁ)-11 with [a]²_D⁵¼ ꢁ231. Based on rac-11, we obtained

36% of (þ)-11 and 22% of (ꢁ)-11. Via the menthyl acrylate of known absolute

configuration and its reaction with diazo(diphenyl)methane, Walborsky et al. con-

nected (þ)-11 with the (R)-configuration [26].

Due to the easier access of (þ)-(R) acid 11, this enantiomer was reduced by the

method described above to the aldehyde (þ)-(R)-14 which showed [a]²_D⁵¼ þ150.5, and

its melting point (52 – 548) was lower than that of rac-14 (74 – 768).

The Wittig olefinations of (þ)-(R)-2,2-diphenylcyclopropane-1-aldehyde (14) were

carried out as described above for rac-14. According to the notation rules, the olefinic

products have the (S)-configuration for (E)-22 as well as for (Z)-22. The specific

rotations, [a]²_D⁵, in Table 2 reveal a regularity for (þ)-(S,E)-22 and (ꢁ)-(S,Z)-22.

Table 2. Optically Active (S)-1-(2-Arylethenyl)-2,2-diphenylcyclopropanes

Formula No.

Aryl

Wittig reaction,

Yield [%]

M.p. [8]

[a]²_D⁵in Decalin

base system

(E)-22a

(E)-22b

(E)-22c

(E)-22d

(E)-22e

(E)-22f

(E)-22g

(E)-22h

(Z)-22a

(Z)-22b

(Z)-22c

(Z)-22d

(Z)-22e

(Z)-22f

(Z)-22g

(Z)-22h

Ph

PhLi/Et₂O

EtONa/EtOH

PhLi/Et₂O

36

12

21

14

18

35

24

23

57

32

29

26

60

56

48

56

85 – 86

76 – 78

100 – 102

76.5 – 78.5

99 – 101

158 – 150

112 – 114

oil

þ 340

þ 268

þ 334

þ 290

þ 191

þ 304

þ 215

þ 71

4-MeꢁC₆H₄

3-ClꢁC₆H₄

4-MeOꢁC₆H₄

4-NO₂ꢁC₆H₄

4-[(E)-Styryl]ꢁC₆H₄

Naphthalen-1-yl

Anthracen-9-yl

Ph

PhLi/Et₂O

EtONa/EtOH

PhLi/C₆H₆

PhLi/Et₂O

EtONa/EtOH

PhLi/Et₂O

oil

ꢁ 232

ꢁ 346

ꢁ 324

ꢁ 323

ꢁ 361

ꢁ 543

ꢁ 224

ꢁ 87

4-MeꢁC₆H₄

3-ClꢁC₆H₄

54.5 – 56.5

oil

91 – 93

128 – 130

86 – 88

oil

4-MeOꢁC₆H₄

4-NO₂ꢁC₆H₄

4-[(E)-Styryl]ꢁC₆H₄

Naphthalen-1-yl

Anthracen-9-yl

PhLi/Et₂O

EtONa/EtOH

PhLi/C₆H₆

2.3. Racemization Rates of (þ)-(S)-1-[2-(E)-Arylethenyl]-2,2-diphenylcyclopro-

panes. When (þ)-(E)-22a was heated in decalin at 1208, the optical rotation

1

continuously decreased. After 9 h, the rotation was close to zero, but the H-NMR

spectrum (CDCl₃) of the isolated product showed no change, and crystalline rac-(E)-

1366

Helvetica Chimica Acta – Vol. 94 (2011)

22a was isolated in high yield. Thus, the racemization can be studied without noticeable

disturbance by the isomerisation to 3,4,4-triphenylcyclopent-1-ene (23a; cf. Scheme 5).

ln (a_o/a_t) ¼ k_ract ¼ 2 k_inv

t

(1)

The rate constants of racemization were measured polarimetrically and evaluated

by the simple first-order law (Eqn. 1). For at least three half-reaction times (87.5%),

the plots of ln (a_o/a_t) vs. time were linear. Rate constants k_racfor (þ)-(E)-22a were

measured at four temperatures over a range of 298 (Table 3), and double runs were

reproducible within ꢀ 1%. The half-reaction time decreases from 246 (99.58) to

15.6 min (128.78). The activation parameters listed in Table 3 will be compared with

those of ring enlargement in Sect. 2.4.2.

Table 3. Kinetics for the Racemization of (þ)-(S)-2,2-Diphenyl-1-[(E)-2-phenylethenyl]cyclopropane

(22a; 20 – 30 mm) in Decalin (Polarimetry)

T [8]

k_rac[ · 10⁴s^ꢁ1

]

Activation parameters

99.5

109.6

119.4

128.7

0.469, 0.469

1.32, 1.30

3.33, 3.33

7.44, 7.35

Arrhenius

E_A¼ 28.2 ꢀ 0.8 kcal mol^ꢁ1

log A ¼ 12.2 ꢀ 0.4

Eyring

DH⁼¼ 27.4 ꢀ 0.8 kcal mol^ꢁ1

DS⁼¼ ꢁ5.2 ꢀ 2.1 e.u.

DG⁼(1008) ¼ 29.3 ꢀ 1.6 kcal mol^ꢁ1

Table 4 contains k_racvalues for (þ)-(E)-22a at 109.68 in 12 solvents which are

ordered by increasing polarity; the E_Tvalues of Reichardt [27] stretch over a wide

range. The numerical response in k_racis minimal and does not show any correlation with

solvent polarity.

Table 4. Rate Constants for Racemization of (þ)-(S,E)-22a (20 – 30 mm) and Ring Enlargement of rac-

(E)-22a (0.15 – 0.20 mm): Variation of Solvent

Solvent

k_rac[ · 10⁴s^ꢁ1

]

k_rac(rel.)

E_T(30)

k_isom[ · 10⁴s^ꢁ1

]

k_isom(rel.)

at (109.6 ꢀ 0.2)8

at (159.3 ꢀ 0.2)8

Decalin

Cyclohexane

Bu₂O

Benzene

Dioxan

PhCl

ClCH₂CH₂Cl

DMF

DMSO

MeCN

BuOH

EtOH

1.31

1.45

1.41

1.74

1.12

1.53

1.18

1.20

1.00

1.13

1.22

1.24

ꢂ1.00

1.11

1.08

1.33

0.86

1.17

0.90

0.92

0.76

0.86

0.93

0.95

31.2

30.9

33.4

34.5

36.0

37.5

41.9

43.8

45.0

46.0

50.2

51.9

56.3

0.932

1.07

0.994

1.18

0.888

1.04

ꢂ1.00

1.15

1.07

1.26

0.95

1.12

0.97

0.900

0.907

a

)

0.964

0.941

1.08

1.03

1.01

1.16

0.88

a

HOCH₂CH₂OH

)

0.821

^a) No measurements.

Helvetica Chimica Acta – Vol. 94 (2011)

1367

Although the arylethenyl groups of (þ)-(E)-22a – 22h harbor electron-attracting

and electron-releasing substituents, the data for k_rac(119.48 in decalin) in Table 5 reveal

an astonishingly small influence of aryl variation. We conclude that the activation

process is not connected with a notable change of charge distribution. The highest rate

increase – still modest – is shown by Ar¼ anthracen-9-yl (k_rac(rel.) ¼ 2.38) which may

indicate radical stabilization.

Table 5. Kinetics of Racemization of (þ)-(S)-1-[(E)-2-Arylethenyl]-2,2-diphenylcyclopropanes (22;

20 – 30 mm) and Ring Enlargement of rac-22 (0.15 – 0.20 mm) in Decalin

Formula No. Ar

k_rac[ · 10⁴s^ꢁ1

at 119.48

]

k_rac(rel.) l [nm] k_isom[ · 10⁴s^ꢁ1

at 159.38

]

k_isom(rel.)

22a

22b

22c

22d

22e

22f

22g

22h

Ph

3.33

3.34

2.73

3.52

4.66

ꢂ 1.00

1.00

0.82

1.06

1.40

1.97

1.39

2.38

269

274

329

350

320

259

0.932

1.01

0.863

1.12

4.45

2.13

2.12

2.11

ꢂ 1.00

1.09

0.93

1.20

4.80

2.28

2.26

4-MeꢁC₆H₄

3-ClꢁC₆H₄

4-MeOꢁC₆H₄

4-NO₂ꢁC₆H₄

4-[(E)-Styryl]ꢁC₆H₄6.57

Naphthalen-1-yl

Anthracen-9-yl

4.64

7.93

2.4. 3-Aryl-4,4-diphenylcyclopent-1-enes (23a – 23h; cf. Scheme 5). 2.4.1. Thermal

Isomerization of (E)-22a – 22h. Without solvent, samples of (E)-22 were heated at 2008

under N₂for 1.5 h, usually followed by high-vacuum distillation and recrystallization, to

furnish the cyclopentenes 23a – 23h (Table 6) in isolated yields of 90 – 96%. The ring

enlargement of the parent compound (E)-22a was described in the preceding paper,

and 23a was characterized by epoxidation to give 25 and by catalytic hydrogenation

[18].

A striking phenomenon is the identity of the mass spectra of styryl-cyclopropanes

22 and the cyclopentenes 23. It was suggested in the preceding paper [18] that the

radical cations of 22 and 23 furnish one and the same open-chain 24 (Scheme 5). This

Table 6. Properties of 3-Aryl-4,4-diphenylcyclopent-1-enes (23)

Formula Aryl

No.

M.p. [8]

d(H) [ppm]

²J(5a,5b)

[Hz]

HꢁC(1) HꢁC(2) HꢁC(3) H_aꢁC(5) H_bꢁC(5)

23a

23b

23c

23d

23e

23f

23g

23h

Ph

68 – 70

59 – 61

oil

99 – 101 5.96

84 – 86 5.97

115 – 117 5.98

6.00

5.97

6.02

5.99

4.78

4.75

4.73

4.74

4.85

4.78

5.62

6.41

2.78

2.79

2.69

2.77

2.81

2.80

2.83

3.09

3.62

3.61

3.60

3.59

3.63

3.78

4.25

16.3

16.2

16^a)

16.3

16.5

16.0

17.1

4-MeꢁC₆H₄

3-ClꢁC₆H₄

4-MeOꢁC₆H₄

4-NO₂ꢁC₆H₄

4-StyrylꢁC₆H₄

5.98^a)

5.99

6.10

6.03

6.10

6.14

Naphthalen-1-yl 135 – 136 6.04

Anthracen-9-yl 135 – 137 6.14

^a) Not sufficiently resolved.

1368

Helvetica Chimica Acta – Vol. 94 (2011)

Scheme 5

more favorable structure is a distonic radical ion in which charge and electron spin are

formally separated [28]. All the fragmentation pathways a – f were observed.

1

The H-NMR spectra of 23a – 23h (Table 6) disclose a high aptitude for H,H

coupling: each of the five aliphatic H-atoms couples with the other four. Apart from

³J(1,2) ¼ 5.9 and ²J(5a,5b) ¼ ꢁ16.3 Hz, the three vicinal, three allylic, and two

homoallylic couplings were shown to be in the range of 1.7 – 2.4 Hz by the computer

simulation of the complex spectrum of 23a; the homoallylic ⁵J(3,5) ¼ 1.7 Hz is even the

same for cis- and trans-relation [18][29].

The cyclopentenes 23 are expected to be in an envelope conformation with C(4) as

the flap. The paramagnetic influence of 3-aryl shifts the cis-located H_bꢁC(5) by 0.8 –

0.9 ppm to lower field, compared with H_aꢁC(5) in trans-position. The two HꢁC(5)

appear as doublets of quadruplets. These are pseudo-quadruplets generated by the

3

4

similarity of J(1,5), J(2,5), and ⁵J(3,5) (cis and trans).

The chemical shifts of the aliphatic H-atoms – the vinylic HꢁC(1) and HꢁC(2) with

³J(1,2) ¼ 5.8 – 5.9 Hz included – do not change much, as long as 3-aryl is a

monosubstituted phenyl group (i.e., 23a – 23f), but the d(HꢁC(3)) shoots up for

Ar¼ naphthalen-1-yl and anthracen-9-yl, and that of 23h passes the signal of the vinylic

H-atoms. One side-ring of anthracenyl is located above the cis-4-phenyl in parallel

planes. A mutual shielding spreads the signals of cis-4-Ph (2 H :1 H :2 H) and one

anthracenyl-H upfield to 6.53 – 7.08 ppm, whereas three anthracenyl-H (out of HꢁC(1’)

to HꢁC(4’)) are shifted up into the region of the ꢄunaffectedꢂ H-atoms of trans-4-

phenyl. Another set of anthracenyl H-atoms, five 1-H signals for HꢁC(5’) to HꢁC(8’)

and the s of HꢁC(10’) remain in the low-field region (7.74 – 8.31), as expected for

anthracene derivatives. Thus, the anthracenyl residue has lost the magnetic equivalence

of the two benzo rings in 23h. A hindrance to rotation about the bond RC₆H₄ꢁC(3) in

1

23a – 23f is not discernible by H- and ¹³C-NMR.

2.4.2. Rate Constants of Ring Enlargement. The styrene-type conjugation in (E)-22a

is lost in the cyclopentene 23a. As a consequence, the UV absorption coefficient in

decalin at 269 nm is diminished by a factor of 40 in the rearrangement. Thus, UV

spectrophotometry was chosen to determine the rate of conversion. The first-order

Eqn. 2 takes the experimentally determined A₁into account; A₁is usually somewhat

larger than the precalculated absorbance of product 23.

Helvetica Chimica Acta – Vol. 94 (2011)

1369

A_oꢁ A₁

ln

¼ k_isom

t

ð2Þ

A_tꢁ A₁

The absorbance A_twas measured up to conversions of 85 – 90%, and the rate

followed the first-order law. The values of k_isomof 22a were measured over a

temperature range of 258, and linear regression provided the activation parameters

(Table 7). The barrier to ring enlargement, DH⁼_isom¼ 31.0 kcal mol^ꢁ1, is by 3.6 kcal

mol^ꢁ1higher than that of racemization. At 1508, a temperature between the ranges of

measurement, (þ)-22a racemizes with t_1/2¼ 2.7 min and rearranges with t_1/2¼ 4.7 h, i.e.,

slower by a factor of 106.

Table 7. Kinetics for the Ring Enlargement of rac-2,2-Diphenyl-1-[(E)-2-phenylethenyl]cyclopropane

(22a; 0.15 – 0.20 mm) in Decalin (UV Spectrophotometry)

T [8]

k_isom[ · 10⁴s^ꢁ1

]

Activation parameters

159.3

169.1

177.5

184.0

0.937, 0.925

2.12, 2.12

4.29, 4.35

6.76, 6.85

Arrhenius

E_A¼ 31.9 ꢀ 1.2 kcal mol^ꢁ1

log A ¼ 12.1 ꢀ 0.6

Eyring

DH⁼¼ 31.0 ꢀ 1.2 kcal mol^ꢁ1

DS⁼¼ ꢁ6.0 ꢀ 2.6 e.u.

DG⁼(1008) ¼ 33.2 ꢀ 2.2 kcal mol^ꢁ1

The influence of the solvent on the rate of rearrangement is as small as that on the

racemization. On comparing the rate constants k_racand k_isomin Table 4, the different

temperatures, 109.68 vs. 159.38, should be kept in mind. Columns 3 and 6 with the

relative values of k_racand k_isom, based on those of 22a, reveal a certain parallelism, but

none whatsoever with E_T, the parameter of solvent polarity. A noteworthy observation:

racemization and ring enlargement smoothly proceed in alcohols. No interaction at

1608 was observed in a preparative experiment with 22a in ethane-1,2-diol, in which 23a

was isolated in high yield.

Somewhat larger, but still unimpressive, is the influence of the aryl substituent on

the rate constant in (E)-22a – 22h. For the photometric rate measurements, a

wavelength was used which offers a great difference of absorbance in (E)-22 and 23

(decalin), not necessarily l_max. The values and the rate constants k_isomare listed in

Table 5; the relative rate constants (columns 4 and 7) on the basis of 22a disclose the

same small magnitude of substituent effects for racemization and rearrangement; with

k_isom(rel.) ¼ 4.80, 22e (Ar¼ 4-NO₂ꢁC₆H₄) is at the top. Further discussion is presented

in Sect. 3.3.

2.5. X-Ray Analyses. The X-ray diffraction pattern reveals the s-trans conformation

of 2,2-diphenyl-1-[(E)-styryl]cyclopropane (22a; Fig. 1). A precise ꢄbisecticꢂ attach-

ment of the styryl residue to the cyclopropane would require orthogonality of the plane

C(1)ꢁC(4)ꢁC(5) and the three-membered ring; the deviation amounts to 5.88. Even

the planarity of the styryl residue is not perfect; the Ph group is twisted vs. the ethylenic

plane by 6.58. Presumably, lattice forces are responsible for these deviations. The bond

lengths of C(1)ꢁC(4) (1.473 ꢅ) and C(4)ꢁC(5) (1.336 ꢅ) differ only on the third

1370

Helvetica Chimica Acta – Vol. 94 (2011)

Fig. 1. ORTEP View of 22a with thermal ellipsoids drawn at 50% probability level

decimal from those of the parent vinylcyclopropane (electron diffraction in gas [30]).

Further features of the X-ray structure will be discussed in connection with calculations

(cf. [19]).

The X-ray diffraction discloses for the 3,4,4-triphenylcyclopent-1-ene (23a) a rigidly

frozen envelope conformation in two crystallographically independent molecules with

very close geometrical parameters; Fig. 2 shows the molecular structure of the first. In

accord with the dihedral angle for C(3)C(2)¼C(1)C(5) of 0.838, the planarity of the

ethylenic bond system is nearly perfect in the crystal. The two edges of the flap,

C(3)ꢁC(4) and C(4)ꢁC(5), show almost identical dihedral angles, ꢁ 32.59(10)8 and

32.65(10)8, undisturbed by 3-Ph group. The flap, C(4), rises from the mentioned

ethylenic plane by 0.56 ꢅ.

3. Discussion. – 3.1. Conformational Control of Racemization and Ring Enlarge-

ment. We have anticipated exo,endo-isomeric open-chain allylic structures 1 and 3.

Both have the capacity to close the three-membered ring, but only the endo-substituted

allylic diradical 3 is capable of 1,5-cyclization. 1,5-Radical recombination of 1 should

lead to a highly strained (unknown) trans-cyclopentene; on the other hand, a rotation

about the C¼C bond should increase the energy balance of activation by an additional

14 kcal mol^ꢁ1for the transient loss of allylic resonance. These considerations are

consistent with diradicals 1 and 3 as true intermediates or as ꢄintermediate phasesꢂ. We

discussed the latter term to describe significant conformations that are neither TS nor

bottom of the energy trough. However, we learned that Northrop and Houk [31] in

2005 had proposed ꢄparaintermediatesꢂ for such species occurring on flat hypersurfaces;

we gladly adopt this designation.

Helvetica Chimica Acta – Vol. 94 (2011)

1371

Fig. 2. ORTEP View of 23a with thermal ellipsoids drawn at 50% probability level

Rotation about the 1,1’-bond of vinylcyclopropane allows several favored

conformations. The bisectic s-trans as the most stable form has found general consent.

The s-cis-conformation may suffer from front-side strain. In 1966, two groups proposed

a mixture of s-trans and s-gauche in a three-well torsional potential with respect to the

1,1’-bond, based on ¹H-NMR phenomena [32]. de Meijere and Lꢀttke reported in 1969

on the electron diffraction of gaseous vinylcyclopropane and established an equilibrium

of 75 ꢀ 6% s-trans and 25 ꢀ 6% s-gauche conformation, corresponding to DG₂₉₃

¼

ꢁ1.06 kcal mol^ꢁ1[30]. With the 1,1’-bond as a vertical axis, the Newman projection

in Scheme 6, with R ¼ H, shows the three conformations on the first line; there was no

evidence for a s-cis-conformation. Our quantum-chemical calculations confirmed the

relative stabilities [19].

The s-trans conformation of 2 appears to be predestined to generate the exo-

substituted diradical 1 on ring-opening. Highly probable as well, the s-gauche-

conformation of 2 is the precursor of the endo-allylic diradical 3. The projections on the

lower line of Scheme 6 serve as illustrations.

On changing from vinylcyclopropane (2) to its triphenyl derivative 22a, the steric

interaction of the (E)-styryl group with the cyclopropane ring remains unaltered

(Scheme 6; R ¼ Ph). The cis-2-Ph group, however, generates strain in one s-gauche

conformation (1208). Whether a true intermediate or a paraintermediate, a con-

formation of this kind must be passed in the 1,2-ring opening leading to the endo-

substituted diradical, which yields the cyclopentene 23a on 1,5-recombination. The Ph

group of the (E)-styryl group winds up in the allylic exo-1-position of diradical 27.

1372

Helvetica Chimica Acta – Vol. 94 (2011)

Scheme 6

A flow-sheet combines enantiomerization and ring enlargement (Scheme 7). It

begins in the upper left with the reversible ring-opening of s-trans-(R)-22a. In the

racemization, the rotational barrier for (Re)-exo,exo-26 to the (Si)-enantiomer is

overcome. It remains hidden, how often (Re)-26 recyclizes, before it undergoes rotation

about the C(3)ꢁC(4) bond.

The 1-exo-Ph group contributes 8.8 kcal mol^ꢁ1to the allyllic stabilization of exo,exo-

26. According to calculations (B3LYP/6-31G*) of the planar 3-exo-methyl-1-exo-

phenylallyl radical as a model, þ 8.8 kcal mol^ꢁ1must be invested for twisting the Ph

group to the orthogonal position. In contrast, a 1-endo-Ph group in the isomeric

structure would add only 4.2 kcal mol^ꢁ1to the conjugation energy of the allylic radical;

the collision with the 3-endo-H requires a 198 rotation about the C(1)ꢁPh bond in the

ground state³).

The less-populated s-gauche conformation of (R)-22a at the lower left of Scheme 7

generates the more favored open-chain structure, (Re)-exo,endo-diradical 27, on

homolysis of the 1,2-bond. Here, rotation to (Si)-exo,endo-27 competes with the 1,5-

ring closure to give cyclopentene (S)-23a, in addition to the continued 3,5-

recombination.

3.2. Rate Increase on Substitution by Three Phenyl Groups. The best activation data

for the thermal rearrangement of the parent vinylcyclopropane are based on the gas

kinetic measurements by Lewis et al. [4]: E_A¼ 51.7 ꢀ 0.5 kcal mol^ꢁ1and log A ¼ 14.3 ꢀ

0.1 s^ꢁ1. Our measurements (Table 7) with 22a in decalin as solvent furnished E_A¼

31.9 kcal mol^ꢁ1and log A ¼ 12.1 s^ꢁ1. The lowering of E_Aby 19.8 kcal mol^ꢁ1reflects

the effect of three Ph groups on the TS of 1,5-cyclization minus their influence in the

ground-state of 22a. Several phenomena are contributing:

1) The diphenylmethyl radical is stabilized by ꢁ 20.4 kcal mol^ꢁ1. Zipse and co-

workers calculated by (G3(MP2)-RAD//UB3LYP/6-31G(d)) an isodesmic H-abstrac-

tion by the CH₃radical [33]; the energy includes the gain by relieving the steric

3

)

We thank W. Sicking, University of Duisburg-Essen, for the calculations.

Helvetica Chimica Acta – Vol. 94 (2011)

1373

1374

Helvetica Chimica Acta – Vol. 94 (2011)

crowding in the ground-state of Ph₂CHꢁ. A slightly higher stabilization is expected for

the diphenylethyl in diradical 27.

2) The conjugation energy of the Ph in the styryl group of 22a may be worth

ꢁ 3 kcal mol^ꢁ1, but the stabilization of this Ph in the diradical exo,endo-27 is higher.

From the model calculations above, we derived þ 8.8 kcal mol^ꢁ1for twisting such an

exo-Ph out of coplanarity. The stabilization may account for ca. ꢁ 6 kcal mol^ꢁ1.

3) In the conformational equilibrium of 22a, s-gauche is less populated than s-trans,

and, on the energy level of the ring-opened diradicals, exo,endo-27 is less favored than

exo,exo-26. In model calculations (UB3LYP/6-31G*) with the 3-methyl-1-exo-phenyl-

allyl radical, a 3-exo-Me group is by 1.1 kcal mol^ꢁ1better than a 3-endo-Me³).

Admittedly, the steric interference of a 2,2-diphenylethyl group on the endo-side of the

allyl radical in 27 exceeds that of Me. A partial cancellation of effects is expected.

4) A certain energy-demanding bending of the planar exo,endo-27 is required to

reach the TS of 1,5-cyclization. This last contribution recalls that we are still dealing

with a TS which does not fully profit from the stabilization of a planar allyl radical.

3.3. Variation of Aryl in the 2-Arylethenyl System. The small influence of aryl

variation on k_racand k_isomof (E)-22 was described above (Sect. 2.3 and 2.4.2, and

Table 5). Substituents in the phenyl group change k_racwithin a range of 2.4, and the

exchange of phenyl by naphthalen-1-yl and anthracen-9-yl is rewarded by a 1.4- and 2.4-

fold increase of k_rac(rel.), respectively.

The minor resonance structure 28 for a segment of (E)-22 resembles the

intermediate 29, which occurs in the radical phenylation of aromatic compounds

(Scheme 8) and furnishes PhꢁAr on H-abstraction. Phenyl radicals have been reacted

in situ with binary mixtures of ArH/benzene, and the product analysis provides ꢄpartial

rate factorsꢂ (p.r.f.), based on the phenylation of one benzene-CH (cf. the review of Hey

[34]). Despite the small range of k_rac, a fairly straight line resulted, when log k_racwas

plotted vs. log p.r.f. of radical phenylation (Fig. 3). The slope leaves no doubt that

aromatic phenylation responds stronger to the variation of substituent R than the

racemization of (þ)-22. Reliable values of p.r.f. were chosen which are based on ¹⁴C

isotope dilution technique [35][36] or on GLC analysis [37].

Scheme 8

The substituent effects on k_racand k_isomof 22a – 22f are ꢄdilutedꢂ by the benzene

ring. Generally, carbon radicals are stabilized by electron-donating and -withdrawing

substituents [33]: on introduction of a MeO group, the methyl radical is stabilized by

7.4 kcal mol^ꢁ1. The stabilization of the radical H₂CꢁCN is likewise 7.4 kcal mol^ꢁ1. 2-

Donor-substituted vinylcyclopropanes rearrange faster than the parent, as shown by

kinetic measurements. When we accept E_A¼ 51.7 kcal mol^ꢁ1for the ring enlargement

of the parent, then a 2-methoxy or 2-dimethylamino group diminish E_Aby 13.0 and

20.6 kcal mol^ꢁ1, respectively [38][39].

Helvetica Chimica Acta – Vol. 94 (2011)

1375

Fig. 3. Relation of log k_rac(rel.) in decalin at 119.48 with the log of the partial rate factors of aromatic

phenylation

Donor- and acceptor-substituted vinylcyclopropanes of type 30 were studied by

Buchert and Reissig (Scheme 9) [40]. The equilibration of the four conceivable (and

identified) stereoisomers of 30 proceeded at 1908 in PhCN much faster than in decalin.

An exo,exo-substituted allylic zwitterion 31 evidences the solvent influence. The

authors recognized the description as zwitterion or diradical as semantic; trimethylene

zwitterions and diradicals are regarded as extremes on a continuous scale. At longer

reaction times, 30 was converted to the four diastereomeric cyclopentenes, likewise

identified, now formed via the exo,endo-allylic intermediate 32.

Scheme 9

3.4. A Heteroanalog of Vinylcyclopropane (E)-22. The rearrangement of vinyl-

cyclopropane is a prototype. Nearly each C-atom can be formally replaced by a

heteroatom, and still the rearrangement works. Formal exchange of HꢁC(1’) of 22 by a

N-atom gives rise to azomethines 33 (Scheme 10). These have been obtained optically

active; their enantiomerization and rearrangement to 4,5-dihydro-3H-pyrroles 35 have

1376

Helvetica Chimica Acta – Vol. 94 (2011)

Scheme 10

been investigated in scope, kinetics, and mechanism in the Munich laboratory [41]. The

corresponding aza-allyl-diradicals 34 in exo,exo- and exo,endo-configuration were

postulated as intermediates.

A surprise: the rate constants of racemization and ring enlargement of the

azomethines 33 and the all-C-systems 22 are numerically fairly similar (Table 8). Even

different solvents and the availability of activation parameters only for 33, Ar¼

4-MeOꢁC₆H₄, did not destroy the close analogy with 22a. The rate constants for both

systems were extrapolated to 1508 and expressed in the same dimension: for the

azomethine, the racemization is 41 times faster than the formation of the 4,5-dihydro-

3H-pyrrole, whereas k_rac/k_isom¼ 106 was observed for 22a.

Table 8. Kinetic Data for N-(4-Methoxybenzylidene)-2,2-diphenylcyclopropylamine (33; Ar¼ 4-

MeOꢁC₆H₄) in PhCN and 2,2-Diphenyl-1-[(E)-styryl]cyclopropane (22a) in Decalin

33 (Ar¼ 4-MeOꢁC₆H₄)

22a

Racemization

k_rac[ · 10⁵s^ꢁ1] at 1508

390

440

DH⁼[kcal mol^ꢁ1

DS⁼[e.u.]

]

26.8 ꢀ 0.8

27.4 ꢀ 0.8

ꢁ 7 ꢀ 2

ꢁ 5.2 ꢀ 2.1

Ring expansion

k_isom[ · 10⁵s^ꢁ1] at 1508

9.46

4.14

DH⁼[kcal mol^ꢁ1

DS⁼[e.u.]

]

33.2 ꢀ 0.9

31.0 ꢀ 1.2

1 ꢀ 2

ꢁ 6.0 ꢀ 2.6

In the Hꢀckel treatment of the allylic p-system, the SOMO (y₂) harbors the single

electron, and the wavefunction has a node in the middle, i.e., the introduction of the N-

atom does not perturb the allylic system.

3.5. Do Radical Recombinations Require Activation? The combination of two H-

atoms to give H₂does not require an activation energy; part of the bond energy must be

transferred to a ꢄthird bodyꢂ. When the recombination of polyatomic radicals involves

drastic conformational changes, the occurrence of an entropic and/or enthalpic barrier

cannot be ruled out a priori. However, the dimerization of primary, sec-, or tert-alkyl

radicals appears to be diffusion-controlled. The capturing of the oct-1-yl radical by the

persistent radical TEMPO (36) at 208 is slower by one order than diffusion rate and

shows E_A¼ 1.8 ꢀ 0.9 kcal mol^ꢁ1[42]. When sterically demanding substituents are

introduced in ethane, the central bond is widened and loses energy, as systematic

studies by Rꢀchardt et al. revealed [43]. Whether the recombination of the strained

radicals has to overcome an activation barrier could not be established with certainty.

Helvetica Chimica Acta – Vol. 94 (2011)

1377

With increase of steric strain, radicals lose the capacity of dimerization and find other

outlets [44][45].

How does resonance stabilization influence the combination of alkyls radicals?

Sustmann and co-workers found E_A¼ 2.8 kcal mol^ꢁ1for the combination of two allyl

radicals in time-resolved measurements (ꢁ 388 to 1198) in solution. However, the rate

constants (0.43 – 9.3 · 10⁹m^ꢁ1s^ꢁ1) still suggest diffusion control [6]. The meaning of

modest activation barriers in diffusion-ruled reactions is unclear.

Beyond all doubts is the dimerization of triphenylmethyl (37), the veteran of

radical chemistry [46], to give 1-(diphenylmethylidene)-4-(triphenylmethyl)cyclohexa-

2,5-diene (38). This structure for the dimer resulted from re-examination in 1968 [47],

but the older measurements of dissociation rates of ꢄhexaphenylethaneꢂ pertain as well

to 38 (Scheme 11). The rate constants, k₁, were obtained from the uptake of NO by

triphenylmethyl at temperatures between ꢁ 208 to þ 108; Ziegler et al. observed E_A¼

20.0 kcal mol^ꢁ1in toluene as solvent [48]. Supplementary measurements at 358 to 608

´

by Krꢁstjansdottir et al. [49] were carried out with a stopped-flow dilution technique and

provided the Eyring parameters, DH⁼₁¼ 20.5 ꢀ 0.2 kcal mol^ꢁ1and DS⁼₁¼ 4.7 ꢀ 0.6, i.e.,

corresponding to a dissociation free energy of DG⁼₁(208) ¼ þ19.1 kcal mol^ꢁ1. The early

spectrophotometric measurements of the equilibrium constant by Ziegler and Ewald

[50] likewise agreed well with later data collected by Lewis and co-workers [51], and

furnished DH_rxn¼ ꢁ10.5 kcal mol^ꢁ1and DS_rxn¼ ꢁ12.0 e.u. in toluene. The dimerization

free energy amounts to ꢁ 14.0 kcal mol^ꢁ1at 208 and leaves DG₂⁼¼ þ5.1 kcal mol^ꢁ1for

the activation free energy of the radical recombination; k₂¼ 1.4 · 10³m^ꢁ1s^ꢁ1(toluene,

208) results for the rate constant of dimerization.

Scheme 11

In the dimerization of triphenylmethyl, the resonance stabilization has to be

relinquished, and the pyramidalization increases the steric strain. Probably, entropic

and enthalpic factors contribute to the activation barrier. The 3,3’,5,5’-tetra(tert-butyl)

derivative of ꢄhexaphenylethaneꢂ is colorless, and the establishing of the dissociation

equilibrium with the yellow monomer requires weeks [52]; a much higher activation

barrier is expected for the dimerization.

3.6. Concerted or Stepwise? All evidence for concertedness is indirect, whereas two-

step processes can be established by interception of intermediates. The diradicals 26

and 27, discussed as possible intermediates, are related to the trimethylene diradical,

i.e., the ring-opened cyclopropane. The failure of intercepting trimethylene diradicals is

usually ascribed to the close vicinity of the internal trap: it is the second radical which

waits for recombination.

The stabilization energy of trimethylenes can be increased by push-pull substitu-

tion. Cram and co-workers observed that the optically active cyclopropane derivative

1378

Helvetica Chimica Acta – Vol. 94 (2011)

40 racemizes with a half-life of 133 min in MeOH at 1258 [53]. The further reaction at

1508 corresponds to a carbocationic S_n1-type behavior, indicating an equilibrium with

the zwitterion 39 (Scheme 12). The energy well of this intermediate is deep enough for

interception by MeOH [54].

Scheme 12

When the ring-opening of cyclopropane involves only one stretching and one

deformation mode, a 90,90-conformation 41 (Scheme 12) will be formed, as defined by

Roald Hoffmann (1968) in a classic paper on trimethylene [55]. The energy of 41 will

rise with increasing angle CꢁCꢁC, and an intermediate is not involved. By two 908

rotations about the CꢁC bonds, the 0,0-conformation 42 is reached; EH-calculations

revealed an intermediate which has to overcome a barrier of 1 kcal mol^ꢁ1for ring

closure. In later ab initio calculations, this barrier disappeared. Numerous theoretical

studies of the electronic configurations of trimethylene, including reaction dynamics,

are of appalling complexity and diversity.

The corresponding ring-opening of vinylcyclopropane reaches a 90,90-conforma-

tion that would energetically not profit from the allyl resonance. Two CꢁC rotations

must accompany the cleavage of the C(1)ꢁC(2) bond to reach the 0,0-conformation,

i.e., the planar allylic system in the diradicals 1 and 3. It is relevant that a 0,90-

trimethylene occurs in diastereomerization, and a 90,90-trimethylene is passed in a one-

step enantiomerization. The substantial stabilization, which the allylic diradicals 26 and

27 experience by conjugation with three Ph groups, may well promote the formation of

a trough in the energy profile. Concerning intermediates or paraintermediates, our

calculations promote a noteworthy dual answer [19].

The Fonds der Chemischen Industrie, Frankfurt, supported our work; we express our thanks. R. H. is

highly indebted to Prof. Ken Houk, UCLA, for an inspiring discussion, and to Prof. Rudolf Knorr,

Munich, for NMR advice. We thank Helmut Schulz for carrying out numerous elemental analyses.

Experimental Part

1. General. See [18].

2. Racemic 1-[2-Arylethenyl]-2,2-diphenylcyclopropanes ( ¼ 1,1’-[2-(2-Arylethenyl)cyclopropane-

1,1-diyl]dibenzenes; 22). 2.1. 2,2-Diphenylcyclopropanecarboxylic Acid (rac-11). Benzophenone hydra-

zone (130 g, 0.67 mol) was oxidized to diazo(diphenyl)methane with HgO, carefully washed alkali-free,

Helvetica Chimica Acta – Vol. 94 (2011)

1379

in Et₂O [56]. After removal of the solvent, the residue was dissolved in light petroleum (40 – 808; 1 l) and

added to ethyl acrylate (180 g, 1.80 mol) with stirring. After 14 h at r.t., the solvent and the excess of

acrylate were distilled at normal pressure, and rac-10 (146 g, 88%) was obtained at 1588/0.5 Torr as a

yellow viscous oil. Ester hydrolysis with KOH (60 g) in MeOH (1200 ml) afforded rac-11 (118.0 g, 90%).

Colorless prisms. M.p. 169 – 1718 ([20]: 169 – 1718).

2.2. 2,2-Diphenylcyclopropane-1-carbaldehyde (rac-14). 2.2.1. Path a. Some drops of DMF were

added to 11 (10.0 g, 42.0 mmol) in SOCl₂(50 ml; Merck, zur Synthese) and kept overnight at r.t.

Evaporation left 12 as crystalline residue. The soln. in abs. Et₂O (decanting from undissolved acid

anhydride) was dropwise added with stirring to dry aziridine (1.81 g, 42.0 mmol) and Et₃N (4.24 g,

41.9 mmol) in abs. Et₂O (100 ml), precooled to ꢁ 258. After 2 h at ꢁ 108, Et₃NHCl was filtered, and the

Et₂O soln. of 13 quickly stirred with 0.55m LiAlH₄in Et₂O (23 ml, 12.6 mmol), kept at 08. Stirring at 08

for 2 h and workup with 2n H₂SO₄/Et₂O and distillation at 140 – 1608/0.001 Torr gave crude product; rac-

14 (3.10 g), m.p. 72 – 748, crystallized from Et₂O, and a further fraction (1.45 g, together 49%) was

obtained by TLC (silica gel Merck, PF_254þ366; cyclohexane/Et₂O 7:1). Analytically pure rac-14 came from

Et₂O at low temp. Colorless prisms. M.p. 74 – 768. IR (KBr): 694vs, 704vs, 753vs, 765s (arom. oop (¼out-

of-plane)); 1440s, 1489s (arom. ring vibr.), 1695vs (C¼O). ¹H-NMR (CCl₄): 1.52 – 2.62 (m, HꢁC(1),

CH₂(3)); 6.95 – 7.45 (m, 10 arom. H); 8.57 (d, ²J ¼ 6.5, CHO). Anal. calc. for C₁₆H₁₄O (222.27): C 86.45,

H 6.35; found: C 86.56, H 6.40.

2,4-Dinitrophenylhydrazone of rac-14. Yellow-orange prisms (EtOH/AcOEt). M.p. 186 – 1878

(dec.). Anal. calc. for C₂₂H₁₈N₄O₄(402.40): C 65.66, H 4.51, N 13.92; found: C 65.92, H 4.56, N 14.28.

2.2.2. Path b. Benzophenone hydrazone (39.2 g, 200 mmol) was converted, as described above, to

diazo(diphenyl)methane, which was reacted in benzene (200 ml) with freshly distilled acrolein (12.0 g,

214 mmol). After keeping at r.t. overnight, the N₂evolution ceased, and the dark violet color turned to

yellow. Filtering, removal of solvent, and distillation at 1508/10^ꢁ3Torr gave rac-14 (31.2 g, 70%), which

solidified as wax; from EtOH at low temp., transparent prisms, m.p. 72 – 748, were obtained, in mixed

m.p., IR und ¹H-NMR identical with those of the specimen of Path a.

2.3. Wittig Olefinations. 2.3.1. With (Benzylidene)(triphenyl)phosphorane (21a). Variant A.

(Benzyl)(triphenyl)phosphonium chloride, m.p. 316 – 3188 (317 – 3188 [57]; 7.10 g, 18.2 mmol), suspend-

ed in abs. Et₂O (50 ml), was stirred with 1.38m ethereal PhLi (13.2 ml, 18.2 mmol) for 30 min at r.t., and a

yellow-orange soln. of 21a was obtained. The soln. of 14 (1.00g, 4.50 mmol) in Et₂O (20 ml) was added,

and the mixture was kept at r.t. overnight. On treatment with H₂O, the excess of phosphonium salt

precipitated and was filtered. After workup with 2n H₂SO₄/Et₂O, from Et₂O/light petroleum 1:1

triphenylphosphine oxide crystallized. M.p. 147 – 1508 (pure 1538). The residue of the mother liquor was

subjected to prep. thick-layer chromatography (TLC) on 20 glass plates (20 ꢃ 20 cm) on a 2-mm layer of

silica gel Merck PF_254þ366; ascending development, usually with light petroleum (40 – 808); after

evaporation of the solvent, the development was repeated four times, until the UV fluorescence showed a

satisfactory separation. The scratched-off material was eluted with Et₂O and crystallized from EtOH.

The faster-moving zone (R_f((Z)-22) is always > R_f((E)-22)) provided (Z)-22a (560 mg, 45%) from

EtOH, m.p. 82 – 868. Compound (E)-22a (402 mg, 32%) was isolated from the second zone; m.p. 54 – 578

(60 – 61.58 [18]).

Variant B. EtONa (18 mmol; from 415 mg Na) in abs. EtOH (100 ml) was reacted under N₂with the

phosphonium salt (7.10 g, 18.2 mmol). After addition of 14 (1.00 g, 4.50 mmol) in abs. EtOH (50 ml) and

reaction at r.t. for 14 h, workup with H₂O/benzene (isolation of Ph₃PO from small volume of ether),

separation by TLC provided (Z)-22a (655 mg, 49%) and (E)-22a (475 mg, 36%). UV (decalin) of

(E)-22a: l_max269 nm (log e 4.35).

1-[(Z)-2-Phenylethenyl]-2,2-diphenylcyclopropane ( ¼ 1,1’-{2-[(Z)-2-Phenylethenyl]cyclopropane-

1,1-diyl}dibenzene; (Z)-22a). Colorless prisms. M.p. 88 – 898 (EtOH). UV (decalin): l_max259 (log e

4.23). IR (ATR (¼attenuated total reflexion)): 689vs, 692vs, 753vs, 762s, 773s (arom. H, oop); 1442s,

1449s, 1491s, 1595m (arom. ring vibr.), 1575w (arom. conjug.), 1635w ((Z)-vinylic str.). ¹H-NMR

(400 MHz): 1.59 (dd, H_aꢁC(3)); 1.69 (dd, H_bꢁC(3)); 2.76 (m, higher order, 14 signals resolved, HꢁC(1));

4.94 (dd, HꢁC(1’)); 6.39 (d, HꢁC(2’)); 7.15 – 7.50 (m, 15 arom. H); ²J(3a,3b) ¼ 4.8, ³J(1,3a) ¼ 5.9 (trans),

³J(1,3b) ¼ 8.7 (cis), ³J(1,1’) ¼ 9.8, ³J(1’,2’) ¼ 11.5. ¹³C-NMR (100 MHz): 24.0, 26.9, 37.8 (C(3), C(1), C(2));

125.9, 126.57, 126.64, 128.7, 132.8 (C(1’), C(2’), 3 arom. p-CH); 127.2, 128.27, 128.31, 128.4, 128.8 130.8

1380

Helvetica Chimica Acta – Vol. 94 (2011)

(double signal height, as expected for free Ph rotation, 12 arom. o- and m-CH); 137.7, 141.2, 146.2 (3

arom. C_q). MS: 296 (37, M^þ), 219 (9, [M ꢁ Ph]^þ), 218 (28, C₁₇H₁^þ₄), 205 (100, C₁₆H₁^þ₃), 204 (95), 178 (24,

C₁₄H^þ₁₀, phenanthrene^þ), 165 (40, C₁₃H₉^þ, fluorenyl^þ), 115 (36, C₉H^þ₇, indenyl^þ), 91 (41, C₇H^þ₇,

tropylium^þ), 77 (9, Ph^þ). Anal. calc. for C₂₃H₂₀(296.39): C 93.20, H 6.80; found C 93.23, H 6.91.

2.3.2. With (4-Methylbenzylidene)(triphenyl)phosphorane (21b). 4-Methylbenzyl bromide (21.2 g,

115 mmol) and Ph₃P (73.0 g, 0.28 mol) were refluxed in xylene for 14 h; from little EtOH crystallized (4-

methylbenzyl)(triphenyl)phosphonium bromide (48.3 g, 93%). M.p. 270 – 2728.

Variant B: Deprotonation of the phosphonium salt (4.1 g, 9.2 mmol) with EtONa/EtOH, reaction

with 14 (500 mg, 2.25 mmol) as described above, and TLC afforded (Z)-22b (423 mg, 61%) and (E)-22b

(157 mg, 22%) from EtOH.

Variant C used EtOLi (0.28m) in EtOH for deprotonation; the total yield was 60%, and the (Z)/(E)

ratio was similar to the one before.

1-[(E)-2-p-Tolylethenyl]-2,2-diphenylcyclopropane (¼1-[(E)-2-(2,2-Diphenylcyclopropyl)ethenyl]-

4-methylbenzene; (E)-22b). M.p. 88 – 898 (EtOH). UV (decalin): l_max269 (log e 4.41). IR (ATR): 697vs,

750vs, 762s, 768vs, 803s (Ph, oop); 823, 841 (p-disubst. arom. H, oop), 954s, 963s ((E)-ethylenic oop);

1

1443s, 1490m, 1596m (arom. breath. modes), 1643w ((E)ꢁHC¼CH, str.). H-NMR (400 MHz): 1.63 (t,

H_aꢁC(3)); 1.78 (dd, H_bꢁC(3)); 2.44 (ddd, HꢁC(1)); 2.33 (s, Me); 5.47 (dd, HꢁC(1’)); 6.58 (d, HꢁC(2’));

7.05 – 7.48 (m, 14 arom. H); ²J(3a,3b) ¼ 4.9, ³J(1,3a) ¼ 5.9 (trans), ³J(1,3b) ¼ 8.7 (cis), ³J(1,1’) ¼ 9.6,

³J(1’,2’) ¼ 15.8; the signals of H_aꢁC(3) and HꢁC(1) are pseudo ts, due to similar coupling constants, here

and in many spectra of 22. ¹³C-NMR (100 MHz): 21.5 (Me); 23.3, 31.6, 37.9 (C(3), C(1), C(2)); 126.3,

127.0, 129.5, 131.0 (C(1’), C(2’), 2 arom. p-CH); 126.1, 127.6, 128.7, 128.8, 129.6, 131.5 (double intensity, 12

arom. o- and m-CH); 135.4, 136.9, 141.8, 147.0 (4 arom. C_q). MS (program MASS): 310 (47, M^þ; HR-MS:

310.1713; calc. 310.1716), 295 (9, [M ꢁ Me]^þ, HR 295.148/295.148), 219 (100, C₁₇H₁^þ₅, [M ꢁ p-tolyl]^þ),

204 (69, C₁₆H^þ₁₂), 165 (23, fluoren-9-yl^þ), 115 (25, [indenyl]^þ), 105 (13, xylyl^þ). Anal. calc. for C₂₄H₂₂

(310.42): C 92.86, H 7.14; found: C 92.59, H 7.11.

1-[(Z)-2-p-Tolylethenyl]-2,2-diphenylcyclopropane (¼1-[(Z)-2-(2,2-Diphenylcyclopropyl)ethenyl]-

4-methylbenzene; (Z)-22b). M.p. 62 – 638 (EtOH). IR (ATR): 688s, 697vs, 711m, 727m, 761vs (Ph, oop),

827s, 839s (p-disubst. arom. oop); 1443m, 1449m, 1495s, 1512m, 1597m (br; arom. breath. modes);

1

1640vw ((Z)-CH¼CH, str.). H-NMR (400 MHz): 1.59 (t, H_aꢁC(3)); 1.70 (dd, H_bꢁC(3)); 2.37 (s, Me);

2.76 (ddd, HꢁC(1)); 4.90 (dd, HꢁC(1’)); 6.37 (d, HꢁC(2’)); 7.13 – 7.50 (4m, 14 arom. H); ²J(3a,3b) ¼ 4.8,

³J(1,3a) ¼ 5.6 (trans), ³J(1,3b) ¼ 8.7 (cis), ³J(1,1’) ¼ 9.9, ³J(1’,2’) ¼ 11.5. ¹³C-NMR (100 MHz): 23.3 (Me);

24.5, 27.5, 38.1 (C(3), C(1), C(2)); 126.2, 127.0, 129.0, 132.5 (C(1’), C(2’), 2 arom. p-CH); 127.6, 128.8,

129.1, 129.4, 129.6, 131.3 (12 arom. o- and m-CH); 135.2, 136.7, 141.7, 146.7 (4 arom. C_q). Anal. calc. for

C₂₄H₂₂(310.42): C 92.86, H 7.14; found: C 93.02, H 7.13.

2.3.3. With (3-Chlorobenzylidene)(triphenyl)phosphorane (21c). The phosphonium bromide (90%,

m.p. 304 – 3068), prepared from 3-chlorobenzyl bromide and Ph₃P by 14-h refluxing in benzene, was

deprotonated as described in Variant A and reacted with 14 (4.50 mmol) as described above. The

separation by TLC gave rise to (Z)-22c (670 mg, 45%) from the faster-moving zone and (E)-22c (441 mg,

30%) from the slower-moving zone.

1-[(E)-2-(3-Chlorophenyl)ethenyl]-2,2-diphenylcyclopropane (¼1-Chloro-3-[(E)-2-(2,2-diphenyl-

cyclopropyl)ethenyl]benzene; (E)-22c). M.p. 77 – 798 (EtOH). UV (decalin): l_max269 nm (log e 4.20).

IR (ATR): 853m, 878m, 896m (m-ClC₆H₄, oop), 958vs, 968m ((E)-CH¼CH, oop); 1493s, 1560m, 1591s

(arom. ring vibr.); 1639m ((E)-CH¼CH, str.). ¹H-NMR (400 MHz): 1.62 (dd, H_aꢁC(3)); 1.77 (dd,

H_bꢁC(3)); 2.41 (dt, HꢁC(1)); 5.49 (dd, HꢁC(1’)); 6.50 (d, HꢁC(2’)); 7.00 – 7.44 (m, 14 arom. H);

²J(3a,3b) ¼ 5.1, ³J(1,3a) ¼ 5.6 (trans), ³J(1,3b) ¼ 8.6 (cis), ³J(1,1’) ¼ 9.7, ³J(1’,2’) ¼ 15.8. ¹³C-NMR

(100 MHz): 23.4, 31.4, 38.3 (C(3), C(1), C(2)); 127.9 (C(2’)); 133.4 (C(1’)); 127.1, 128.36, 128.45, 130.9

(double intensity, 8 o- and m-CH of 2 Ph); 134.4, 139.6, 141.2, 146.3 (4 arom. C_q); 123.9, 125.7, 126.0,

126.6, 126.8, 129.6 (2 p-CH of Ph₂, 4 CH of C₆H₄Cl); the assignment of the olefinic C-atoms was based on

a 2D GHSQC experiment. MS: 330 (24, M^þ; HR-MS: 330.1173; calc. for C₂₃H₁₉³⁵Cl^þ, 330.1171), 332 (8.7;

HR-MS: 332.1143; calc. for C₂₃H₁₉³⁷Cl, 332.1141; intensity 7.8/8.7), 239 (31; HR-MS: 239.0630; calc. for

C₁₆H₁₂³⁵Cl^þ, 239.0625; intensity of C₁₆H₁₂³⁷Cl 9.9/11.0), 205 (100, C₁₆H₁^þ₃, [M ꢁ CH₂C₆H₄Cl]^þ), 191 (14;

HR-MS: 191.0844; calc. for C₁₅H^þ₁₁, 191.0858), 165 (22, Fluorenyl^þ; HR-MS: 165.0701; calc. for C₁₃H^þ₉,

Helvetica Chimica Acta – Vol. 94 (2011)

1381

165.0702), 91 (18; HR-MS: 91.0547; calc. for C₇H^þ₇, 91.0546). Anal. calc. for C₂₃H₁₉Cl (330.48): C 83.49, H

5.79; found: C 83.72, H 5.70.

1-[(Z)-2-(3-Chlorophenyl)ethenyl]-2,2-diphenylcyclopropane (¼1-Chloro-3-[(Z)-2-(2,2-diphenyl-

cyclopropyl)ethenyl]benzene; (Z)-22c). M.p. 67 – 688 (EtOH). IR (ATR): 820m, 844m, 893s (3-ClC₆H₄,

oop); 1627w ((Z)-CH¼CH, str.). ¹H-NMR (400 MHz): 1.60 (t, H_aꢁC(3)); 1.71 (dd, H_bꢁC(3)); 2.68 (dt,

HꢁC(1)); 4.99 (dd, HꢁC(1’)); 6.33 (d, HꢁC(2’)); 7.12 – 7.49 (m, 14 arom. H); ²J(3a,3b) ¼ 4.9, ³J(3a,1) ¼

3

6.0 (trans), ³J(3b,1) ¼ 8.7 (cis), J(1,1’) ¼ 10.0, J(1’,2’) ¼ 11.5. ¹³C-NMR (100 MHz, DEPT): 24.2, 27.0,

38.5 (C(3), C(1), C(2)); 126.5 – 134.7 (14 arom. CH þ 2 vinyl-C); 134.6, 139.9, 141.5, 146.3 (4 arom. C_q).

Anal. calc. for C₂₃H₁₉Cl (330.48): C 83.49, H 5.79; found: C 83.86, H 5.77.

2.3.4. With (4-Methoxybenzylidene)(triphenyl)phosphorane (21d). By Variant A (PhLi), the wine-

red phosphorane 21d was set free from (4-methoxybenzyl)(triphenyl)phosphonium bromide (m.p. 235 –

2368 (dec.)) and reacted with 14 (1.00 g, 4.50 mmol). Workup by TLC (20 plates; light petroleum/Et₂O

96 :4) gave (Z)-22d (604 mg, 41%) and (E)-22d (410 mg, 28%).

1-[(E)-2-(4-Methoxyphenyl)ethenyl]-2,2-diphenylcyclopropane (¼1-[(E)-2-(2,2-Diphenylcyclopro-

pyl)ethenyl]-4-methoxybenzene; (E)-22d). M.p. 61 – 638 (EtOH). UV (decalin): l_max274.5 (log e 4.56).

IR (ATR): 807s, 823s, 829s, 879m (C₆H₄, oop); 956vs, 958s ((E)-CH¼CH, oop); 1024s (CꢁOꢁC, sym.

str.); 1239vs, 1247s (CꢁOꢁC, asym. str.); 1575w (arom. conj.), 1639w ((E)-vinylic str.). ¹H-NMR

(400 MHz): 1.59 (t, H_aꢁC(3)); 1.73 (dd, H_bꢁC(3)); 2.40 (dt, HꢁC(1)); 3.78 (s, MeO); 5.34 (dd, HꢁC(1’));

6.52 (d, HꢁC(2’)); 6.78 (d, HꢁC(3/5) of 4-MeOC₆H₄), 7.1 – 7.43 (m, 12 arom. H); ²J(3a,3b) ¼ 5.0,

³J(3a,1) ¼ 5.9 (trans), ³J(3b,1) ¼ 8.6 (cis), ³J(1,1’) ¼ 9.5, ³J(1’,2’) ¼ 15.8. ¹³C-NMR (100 MHz, DEPT):

23.2, 31.5, 37.8 (C(3), C(1), C(2)); 55.7 (MeO); 114.3 (C(3/5) of MeOC₆H₄); 126.2 – 131.5 (9 lines for

arom. CH); 131.0, 141.9, 147.0, 159.0 (4 arom. C_q). MS: 326 (86, M^þ; HR-MS: 326.1657; calc., 326.1665;

¹³C: 22.3; calc., 23.0), 295 (6, [M ꢁ MeO]^þ), 235 (71, C₁₇H₁₅O^þ, ¹³C: 12.8; calc., 13.3), 205 (100, C₁₆H^þ₁₃

,

[M ꢁ CH₂C₆H₄OMe]^þ, 191 (21, C₁₅H₁^þ₁, [1-phenylindenyl]^þ), 165 (28, fluoren-9-yl^þ), 121 (47,

[CH₂C₆H₄OMe]^þ), 91 (31, C₇H^þ₇). Anal. calc. for C₂₄H₂₂O (326.42): C 88.30, H 6.79; found: C 88.07,

H 6.64.

1-[(Z)-2-(4-Methoxyphenyl)ethenyl]-2,2-diphenylcyclopropane (¼1-[(Z)-2-(2,2-Diphenylcyclopro-

pyl)ethenyl]-4-methoxybenzene; (Z)-22d). Colorless prisms. M.p. 83 – 848 (EtOH). IR (ATR): 813m,

834m, 845vs (p-disubst. C₆H₄, oop), 1029vs (br., CꢁOꢁC, sym. str.); 1176s, 1246s, 1258s (CꢁOꢁC, asym.

str.); 1569w (arom. conj.). ¹H-NMR (400 MHz): 1.57 (t, H_aꢁC(3)); 1.69 (dd, H_bꢁC(3)); 2.72 (dt,

HꢁC(1)); 3.82 (s, MeO), 4.85 (dd, HꢁC(1’)); 6.33 (d, HꢁC(2’)); 6.89 (d, HꢁC(3/5) of MeOC₆H₄); 7.2 –

7.45 (m, 12 arom. H); ²J(3a,3b) ¼ 4.8, ³J(3a,1) ¼ 6.0 (trans), ³J(3b,1) ¼ 8.7 (cis), ³J(1,1’) ¼ 9.8, ³J(1’,2’) ¼

11.5. ¹³C-NMR (100 MHz, DEPT): 24.5, 27.5, 38.07 (C(3), C(1), C(2)); 55.7 (MeO); 114.1 (C(3/5) of 4-

MeOC₆H₄); 126.2, 127.0, 128.65, 131.53 (2 arom. p-CH, 2 vinyl C); 127.5, 128.72, 130.4, 131.3 (double

intensity, 8 arom. o- and m-CH); 130.7, 141.7, 146.7, 158.7 (4 arom. C_q). Anal. calc. for C₂₄H₂₂O (326.42): C

88.30, H 6.79; found: C 88.34, H 6.97.

2.3.5. With (4-Nitrobenzylidene)(triphenyl)phosphorane (21e). The phosphonium bromide, m.p.

261 – 2628 (dec.) (2608 [57]), prepared as usual, was converted to the cherry-red 21e by Variant B and

reacted with 14 (1.13 g, 5.1 mmol). PLC (5 ꢃ development, light petroleum/ether 96 :4) provided (Z)-22e

(1.03 g, 59%) and (E)-22e (315 mg, 18%).

1-[(E)-2-(4-Nitrophenyl)ethenyl]-2,2-diphenylcyclopropane (¼1-[(E)-2-(2,2-Diphenylcyclopropyl)-

ethenyl]-4-nitrobenzene; (E)-22e). Pale-yellow needles. M.p. 98 – 1008 (EtOH). UV (decalin): l_max329

(log e 4.23). IR (ATR): 861m (C₆H₄NO₂, oop); 953s, 967m ((E)-CH¼CH, oop); 1335vs, 1507s (NO₂,

sym. and asym. str.); 1640m ((E)-CH¼CH, str.). ¹H-NMR (400 MHz): 1.69 (t, HꢁC(3a)); 1.82 (dd,

HꢁC(3b)); 2.46 (dt, HꢁC(1)); 5.68 (dd, HꢁC(1’)); 6.62 (d, HꢁC(2’)); 7.14 – 7.43 (m, 12 arom. H, 2 vinyl

H); 8.09 (d, 2 m-H of NO₂C₆H₄); ²J(3a,3b) ¼ 5.1, ³J(3a,1) ¼ 5.5 (trans), ³J(3b,1) ¼ 8.6 (cis), ³J(1,1’) ¼ 9.8,

³J(1’,2’) ¼ 15.8. ¹³C-NMR (100 MHz): 23.6, 31.6, 39.0 (C(3), C(1), C(2)); 124.1 – 131.2 (16 arom. CH,

C(1’), C(2’)); 141.4, 144.5, 146.3, 146.7 (4 arom. C_q). Anal. calc. for C₂₃H₁₉NO₂(341.39): C 80.91, H 5.61,

N 4.10; found: C 80.71, H 5.52, N 4.27.

1-[(Z)-2-(4-Nitrophenyl)ethenyl]-2,2-diphenylcyclopropane (¼1-[(Z)-2-(2,2-Diphenylcyclopropyl)-

ethenyl]-4-nitrobenzene; (Z)-22e). M.p. 112 – 1138 (EtOH). IR (ATR): 855s, 866s (O₂NC₆H₄, oop),

960m ((Z)-CH¼CH, oop); 1333vs, 1591s (NO₂, sym. and asym. str.); 1629m ((Z)-CH¼CH, str.).

¹H-NMR (300 MHz): 1.66 (t, HꢁC(3a)); 1.78 (dd, HꢁC(3b)); 2.66 (dt, HꢁC(1)); 5.15 (dd, HꢁC(1’)); 6.41

1382

Helvetica Chimica Acta – Vol. 94 (2011)

(d, HꢁC(2’)); 7.12 – 7.42 (m, 10 arom. H, Ph₂); 7.57 (d, further split, 2 o-H of NO₂C₆H₄); 8.20 (d, further

split, 2 m-H of NO₂C₆H₄); ²J(3a,3b) ¼ 4.8, ³J(3a,1) ¼ 5.6 (trans), ³J(3b,1) ¼ 8.8 (cis), ³J(1,1’) ¼ 10.2,

³J(1’,2’) ¼ 11.6. ¹³C-NMR (75 MHz): 24.1, 27.1, 38.5 (C(3), C(1), C(2)); 126.7 (C(2’)); 137.1 (C(1’)); 123.7

(C(3/5) of NO₂C₆H₄); 129.3 (C(2/6) of NO₂C₆H₄); the assignments are based on an HSQCAD

experiment. Anal. calc. for C₂₃H₁₉NO₂(341.39): C 80.91, H 5.61, N 4.10; found: C 80.84, H 5.58, N 3.94.

2.3.6. With (Triphenyl){4-[(E)-styryl]benzylidene}phosphorane (22f). {4-[(E)-2-Phenylethenyl]-

benzyl}(triphenyl)phosphonium Bromide (16). (E)-4-Bromomethylstilbene [22] (15; 3.84 g, 14.1 mmol)

and Ph₃P (3.66 g, 14.0 mmol) in xylene (100 ml) were refluxed for 14 h. The product crystallized from

EtOH. M.p. 297 – 2998 (dec.). Anal. calc. for C₃₃H₂₈BrP (535.45): C 74.02, H 5.27; found: C 73.95, H 5.38.

Wittig Reaction by Variant B. Compound 16 (9.65 g, 18.0 mmol) was deprotonated by EtONa/EtOH

and reacted with 14 (1.00 g, 4.5 mmol). From the soln. of the crystalline crude product in hot benzene/

EtOH, (Z)-22f (840 mg) was obtained. M.p. 115 – 1238. PLC (light petroleum/benzene 95 :5; 3 ꢃ de-

velopment) furnished further (Z)-22f (182 mg, together 57%) and (E)-22f (580 mg, 32%).

2,2-Diphenyl-1-{2-[(E)-4-styrylphenyl]ethenyl}cyclopropane (¼1-[(E)-2-(2,2-Diphenylcyclopropyl)-

ethenyl]-4-[(E)-2-phenylethenyl]benzene; (E)-22f). M.p. 144 –1458 (EtOH). UV (decalin): l_max337.5

(log e 4.40). IR (ATR): 690vs, 718m, 748vs (Ph, oop), 800s (p-disubst. C₆H₄, oop), 960vs, 953s ((E)-

1

CH¼CH, oop), 1444m, 1492s, 1594m (arom. ring vibr.), 1633 ((E)-CH¼CH, str., broadened). H-NMR

(60 MHz): 1.51–2.70 (m, H₂C(3), HꢁC(1)); 5.48 (dd, ³J(1,1’)¼ 9.2); 6.68 (d, ³J(1’,2’) ¼ 16.0); 6.93–7.19 (m,

19 arom. H þ 2 vinyl H of stilbene group). Anal. calc. for C₃₁H₂₆(398.51): C 93.42, H 6.58; found: C 93.47, H

6.61.

2,2-Diphenyl-1-{(Z)-2-[(E)-4-styrylphenyl]ethenyl}cyclopropane (¼1-[(Z)-2-(2,2-Diphenylcyclo-

propyl)ethenyl]-4-[(E)-2-phenylethenyl]benzene; (Z)-22f). M.p. 125 – 1278 (EtOH). IR (KBr): 818vs,

828m (p-disubst. C₆H₄, oop); 1446s, 1492s, 1596w (arom. ring vibr); 1628w and 1636 (sh, (Z)-CH¼CH,

str.). ¹H-NMR (400 MHz): 1.61 (t, HꢁC(3a)); 1.74 (dd, HꢁC(3b)); 2.79 (m, HꢁC(1)); 4.96 (dd,

HꢁC(1’)); 6.39 (d, HꢁC(2’)); 7.12 (s, 2 vinyl H); 7.06 – 7.60 (m, 19 arom. H); ²J(3a,3b) ¼ 4.8, ³J(3a,1) ¼ 5.6

(trans), ³J(3b,1) ¼ 8.7 (cis), ³J(1,1’) ¼ 10.0, ³J(1’,2’) ¼ 11.5. ¹³C-NMR (DEPT): 24.6, 27.7, 38.3 (C(3), C(1),

C(2)); 136.1, 137.5, 137.8, 141.6, 146.6 (5 arom. C_q); 126.3 – 132.2 (21 lines printed, 19 arom. CH, 4 vinyl

H). Anal. calc. for C₃₁H₂₆(398.51): C 93.42, H 6.58; found: C 93.25, H 6.41.

2.3.7. With [(Naphthalen-1-yl)methylidene](triphenyl)phosphorane (21g). The phosphonium chlor-

ide, m.p. 294 – 2958 (dec.), prepared from 1-(chloromethyl)naphthalene and Ph₃P in refluxing xylene, was

deprotonated by EtONa/EtOH (Variant B) and reacted with 14 (1.00 g, 4.50 mmol). The workup by PLC

(light petroleum/benzene 95 :5, 3 ꢃ development) yielded (Z)-22g (974 mg, 62%) and (E)-22g (306 mg,

20%).

1-[(E)-2-(Naphthalen-1-yl)ethenyl]-2,2-diphenylcylopropane (¼1-[(E)-2-(2,2-Diphenylcyclopropyl)-

ethenyl]naphthalene; (E)-22g). M.p. 108 –1108 (EtOH). UV (decalin): l_max230 (log e 3.82). IR (ATR):

950vs, 962m ((E)-CH¼CH, oop), 1444m, 1493s, 1590m, 1595m (arom. ring vibr.); 1639w ((E)-CH¼CH,

str.). ¹H-NMR (60 MHz): 1.53–1.90 (m, 2 H, HꢁC(3a) and HꢁC(3b) not fully separated); 2.56 (dt,

HꢁC(1)); 5.17 (dd, ³J(1,1’) ¼ 9.3, ³J(1’,2’) ¼ 15.3, HꢁC(1’)); HꢁC(2’) signal not separated from arom. H. MS

(program MASS): 346 (62, M^þ; HR-MS: 346.1717; calc., 346.1716; ¹³C: 19.7; calc., 18.8), 268 (8, [M ꢁ

C₆H₆]^þ, HR-MS: 268.1238; calc. 268.1248), 255 (97, C₂₀H₁^þ₅, [M ꢁ PhCH₂]^þ; HR-MS: 255.1160; calc.,

255.1170), 205 (100, C₁₆H^þ₁₃, [M ꢁ naphthyl-CH₂]^þ; HR-MS: 205.1001; calc., 205.1014), 179 (50, C₁₄H₁^þ₁,

perhaps 9-methylfluoren-9-yl^þ), 167 (20, Ph₂CH^þ), 165 (69, fluoren-9-yl^þ), 115 (25, C₉H^þ₇, indenyl^þ), 91

(24, C₇H^þ₇). Anal. calc. for C₂₇H₂₂(346.45): C 93.60, H 6.40; found: C 93.35, H 6.51.

1-[(Z)-2-(Naphthalen-1-yl)ethenyl]-2,2-diphenylcyclopropane (¼1-[(Z)-2-(2,2-Diphenylcyclopro-

pyl)ethenyl]naphthalene; (Z)-22g). M.p. 107 – 1088 (EtOH). IR (ATR): 1443m, 1491m, 1497m, 1589w,

1596w (arom. ring vibr.), 1635w ((Z)-CH¼CH, str.). ¹H-NMR (400 MHz): 1.57 (dd, HꢁC(3b)); 1.61 (t,

HꢁC(3a)); 2.55 (dt, HꢁC(1)); 5.16 (dd, HꢁC(1’)); 6.86 (d, HꢁC(2’)); 7.05 – 8.12 (m in 9 groups, 17 arom.

H); ²J(3a,3b) ¼ 4.7, ³J(3a,1) ¼ 5.8 (trans), ³J(3b,1) ¼ 8.7 (cis), ³J(1,1’) ¼ 10.3, ³J(1’,2’) ¼ 11.4. ¹³C-NMR

(DEPT): 23.7, 26.9, 38.2 (C(3), C(1), C(2)); 127.8, 128.62, 128.82, 131.1 (lines of double intensity, 8 arom.

o- and m-CH of Ph₂); 132.4, 134.1, 135.2, 142.0, 146.6 (5 arom. C_q); 10 more lines for 11 further arom. CH.

Anal. calc. for C₂₇H₂₂(346.45): C 93.60, H 6.40; found: C 93.78, H 6.51.

2.3.8. With (Anthracen-9-ylmethylidene)(triphenyl)phosphorane (21h). 9-[(Dimethylamino)methyl]-

anthracene (18): Anthracene-9-carbaldehyd (17; 5.00 g, 24.2 mmol), HCOOH (10 g), and DMF (8.0 ml)

Helvetica Chimica Acta – Vol. 94 (2011)

1383

were heated at 150 – 1608 for 5 h, whereby H₂O formed was distilled off. On workup with 20% aq. NaOH,

an oil separated, which was solidified, washed, dried, and distilled at 170 – 1758/10^ꢁ3Torr: 18 (5.10 g,

88%). M.p. 63.5 – 658 (light petroleum). ¹H-NMR (80 MHz): 2.14 (s, Me₂N); 4.05 (s, NCH₂). Anal. calc.

for C₁₇H₁₇N (235.31): C 86.76, H 7.28, N 5.95; found: C 86.69, H 7.17, N 6.31.

9-(Trimethylammoniomethyl)anthracene Iodide (19). MeI (20 mmol) was added to 18 (4.50,

19.1 mmol) in Et₂O (50 ml), and the mixture was kept at r.t. for a week. Filtering and air-drying gave 19

(6.63 g, 92%). Yellow crystal powder. M.p. 185 – 1878 (dec.).

(Anthracen-9-ylmethyl)(triphenyl)phosphonium Iodide (20). Ph₃P (40 mmol) was added to the

suspension of 19 (5.00 g, 13.2 mmol) in BuOH (300 ml). The stirred mixture was heated for 4 d at 135 –

1408; Me₃N was eliminated by N₂, and the salt 20 precipitated on cooling. Washing with EtOH and drying

gave 20 (7.06 g, 92%). Yellow prisms (from EtOH). M.p. 265 – 2678 (dec.). ¹H-NMR (60 MHz,

(D₆)DMSO): 6.13 (d, J(P,H) ¼ 15.0, CH₂). Anal. calc. for C₃₃H₂₆IP (580.42): C 68.28, H 4.52; found: C

68.62, H 4.46.

The Wittig reaction by Variant B with 14 (4.50 mmol) furnished a crude product, which by digestion

with light petroleum left triphenylphosphane oxide (1.11 g, 89%) and (E)-22h (143 mg); the latter, m.p.

130 – 1338, remained undissolved, when the solid was treated with cold EtOH. Slow evaporation of the

light petroleum soln. afforded a second crystalline fraction of (E)-22h (151 mg). The residue of the

mother liquor was subjected to PLC to provide (E)-22h (106 mg; together 22%) and (Z)-22h (1.04 g,

58%). A second experiment by Variant C (PhLi in benzene) produced 65% yield with (Z)/(E) 2.8.

1-[(E)-2-(Anthracen-9-yl)ethenyl]-2,2-diphenylcyclopropane (¼ 9-[(E)-2-(2,2-Diphenylcyclopropyl)-

ethenyl]anthracene; (E)-22h). Yellow glistening crystals. M.p. 135 –1368 (EtOH). UV (decalin): l_max259

(log e 4.89), 385 (3.54). IR (ATR): 966s ((E)-CH¼CH, oop), 1445s, 1493s, 1597w (arom. ring vibr.), 1642

((E)-CH¼CH, str.). ¹H-NMR (400 MHz): 1.77 (dd, H_bꢁC(3)), 1.85 (t, H_aꢁC(3)); 2.82 (dt, HꢁC(1)); 5.47

(dd, HꢁC(1’)); 7.21 –7.40 (m, 10 arom. H of Ph₂, HꢁC(2’)); anthracenyl part (9 H): 7.43, 7.54 (2d, 2 ꢃ 2 H,

HꢁC(3’/8’) and HꢁC(4’/7’)); 7.94, 8.08 (2d, further split, 2ꢃ 2 H, HꢁC(5’/6’) and HꢁC(2’/9’); 8.30 (s,

HꢁC(10’)); ²J(3a,3b) ¼ 5.1, ³J(3a,1) ¼ 5.8 (trans), ³J(3b,1) ¼ 8.6 (cis), ³J(1,1’) ¼ 8.7, ³J(1’,2’) ¼ 16.0.

¹³C-NMR (DEPT): 22.5, 31.7, 38.2 (C(3), C(1), C(1), C(2)); 5 arom. C_q(3 for rotating anthracenyl þ 2

for two Ph); eight lines of ꢄdouble intensityꢂ would be expected for arom. CH, but only 6 can be

distinguished; coincidence may play a role. Anal. calc. for C₃₁H₂₄(396.50): C 93.90, H 6.10; found: C 93.80,

H 6.02.

1-[(Z)-2-(Anthracen-9-yl)ethenyl]-2,2-diphenylcyclopropane (¼ 9-[(Z)-2-(2,2-Diphenylcyclopropyl)-

ethenyl]anthracene; (Z)-22h). Yellow oil. ¹H-NMR (60 MHz): 2.0 –3.0 (m, H_aꢁC(3), H_bꢁC(3), HꢁC(1));

5.47 (dd, ³J(1,1’)¼ 9.5, ³J(1’,2’) ¼ 11.0, HꢁC(1’)); 6.7–8.5 (m, 19 arom. H, HꢁC(2’)).

3. Optically Active 1-(2-Arylethenyl)-2,2-diphenylcyclopropanes (¼1,1’-[2-(2-Arylethenyl)cyclopro-

pane-1,1-diyl]dibenzenes). 3.1. Resolution of 2,2-Diphenylcyclopropane-1-carboxylic Acid (11). Accord-

ing to [25], rac-11 (57.0 g, 0.24 mol) was dissolved in acetone (1 l) and added to a boiling soln. of quinine

(Merck; for resolution; 77.4 g, 0.24 mol) in acetone (3.5 l). Crystallization of the quinine salt from the

cooled soln. required inoculation; a small sample, diluted with light petroleum, provided crystals on

scratching. Within 2 d in the refrigerator, the acetone soln. furnished the quinine salt of (þ)-(R)-11

(57.6 g, 43%). M.p. 185 – 1898. Recrystallization from acetone gave fine colorless needles (47.2 g, 35%)

with m.p. of 189.5 – 1918 and [a]²_D⁵¼ 25.1 (CHCl₃, 5 mg/ml). The stirred suspension of the salt in acetone

(150 ml) just dissolved, when conc. aq. HCl (25 ml) was added. After stirring for 15 min at r.t., mixing

with H₂O (1 l), and keeping for 8 h in the refrigerator, the free acid (þ)-(R)-11 precipitated; 20.7 g. M.p.

149 – 1518 ([20]: 150 – 1518). [a]²_D⁵¼ þ226 (CHCl₃). Recrystallization from light petroleum (80 – 1208)

changed neither m.p. nor [a]²_D⁵notably. The combined acetone mother liquors were evaporated to half of

the volume; the crystalline salt, on treating with aq. HCl, furnished impure (ꢁ)-(S)-11 (26.1 g,

0.105 mol). [a]²_D⁵¼ ꢁ165. Its soln. in acetone (130 ml) was combined with dry brucine (Fluka; 47.8 g,

0.105 mol) in acetone (200 ml) and refluxed for 1 h; after filtering and inoculating, the brucine salt of (ꢁ)-

(S)-11 (34.5 g) crystallized within several weeks at 88. Colorless prisms, [a]²_D⁵¼ ꢁ78 (CHCl₃). Treatment

with aq. HCl yielded (ꢁ)-(S)-11 (12.3 g, 22%). M.p. 150 – 1528. [a]²_D⁵¼ ꢁ231 (CHCl₃).

3.2. (þ)-(R)-2,2-Diphenylcyclopropane-1-carbaldehyde (14). The procedure described for rac-11

was repeated with (þ)-(R)-11 (42.0 mmol). The crude aldehyde was not distilled, but immediately

subjected to TLC (25 plates, silica gel, 2 mm; light petroleum/Et₂O 85 :15). The major zone afforded

1384

Helvetica Chimica Acta – Vol. 94 (2011)

(þ)-(R)-11 (4.05 g, 43%), which crystallized. M.p. 45 – 508. From Et₂O at low temp., small colorless

prisms were obtained. M.p. 52 – 548. [a]_D²⁵¼ þ150.5 (CHCl₃). ¹H-NMR and R_fwere identical with those

of rac-14, but the m.p. is lower.

3.3. Wittig Olefinations with (þ)-(R)-14. The experiments with rac-14, described in Sect. 2.3, were

repeated with (þ)-(R)-14 on the same scale (4.5 mmol). The procedure was not optimized; usually, one

experiment was carried out with each methylidenephosphorane. The identity of rac- and (S)-22 in R_fand

¹H-NMR spectrum was in each case demonstrated. Melting points are different; when a compound (S)-

22 remained oily, a second TLC confirmed the purity. The specific rotations, [a]²_D⁵, in decalin were

measured for specimens which were recrystallized from EtOH, whenever possible. Data are compiled in

Table 2.

3.4. Measurements of Racemization Rates of (þ)-(S,E)-22. The technique may be described with an

example. About 60 mg of (þ)-(E)-22a (29 mm) was dissolved in decalin (8 ml), filtered, and filled with an

injection syringe in 0.6-ml portions into 12 ampoules (1 ml), which were carefully cleaned and dried

before.

After sealing, the ampoules were fixed in a frame and immersed into the preheated polyglycol bath

(119.48) of an ultrathermostat (Colora, HT 13). After 15 min, temp. constancy (ꢀ0.38) was reached, the

first ampoule was taken out, cooled with ice water, and filled into the polarimeter tube (5 cm, 0.5-ml

volume) of a light-electric precision polarimeter (LEP; C. Zeiss). Since first-order reactions can be cut

off from the start side, the first value, a ¼ þ5.728, was defined as a_o. The withdrawal times of ten further

ampoules were so chosen that three half-reaction times were covered; in our example, a_t¼ 0.518 (91%

reaction) was measured after 120 min. The last ampoule served as control for a_e¼ 0.008 after 10 half-

times (here 500 min).

The plot of ln (a_o/a_t) vs. t furnished the first-order rate constant k_racfrom the slope of the straight line

or – more elegantly – by linear regression; the correlation coefficient R is a measure of quality. In the

experiment described for 22a, R ¼ 0.999 was found up to 91% reaction. The activation parameters were

evaluated from measurements at different temperatures (Table 3) by linear regression: the activation

energy E_Aand the action constant A from k_racas a function of 1/T by the Arrhenius equation, and

activation enthalpy and entropy, DH⁼and DS⁼, resp., from k_rac/T as a function of 1/T by the Eyring

equation. Here, too, R ¼ 0.999 was observed. Generous estimates for the errors in the rate constants

(ꢀ 5%) and the temp. (ꢀ0.38) led to the tolerances given in Table 3.

3.6. Thermal Isomerization of Vinylcyclopropanes (E)-22 to 3-Aryl-4,4-diphenylcyclopent-1-enes 23.

3.6.1. 3,4,4-Triphenylcyclopent-1-ene (¼1,1’,1’’-Cyclopent-3-ene-1,1,2-triyltribenzene; 23a). The descrip-

tion in the preceding article [18] is supplemented: a) rac-(E)-22a (2.02 g, 6.82 mmol) was heated in a

sealed ampoule under N₂at 2008 for 90 min; distillation at 1708 (bath)/10^ꢁ3Torr gave 23a (1.83 g, 91%),

1

m.p. 64 – 678; the H-NMR spectrum [18] showed no additional signals. Recrystallization from EtOH

furnished fine leaflets, m.p. 68 – 698 (68 – 708 [18]). b) A sample of (þ)-(E)-22a was subjected to

thermolysis (1.5 h, 2008) and showed a ¼ 0.018 (decalin; 25 mg/ml); the mixed m.p. established the

identity with rac-23a. c) The soln. of rac-(E)-22a (160 mg) in decalin (0.6 ml) was sealed in an NMR tube

and heated at 1808 (thermostat). In intervals of ca. 5 min, the region of olefinic H-atoms was checked; at

sweep width of 100 Hz, only the signals of (E)-22a and 23a were observed.

3.6.2. 3-(4-Methylphenyl)-4,4-diphenylcyclopent-1-ene (¼1-(5,5-Diphenylcyclopent-2-en-1-yl)-4-

methylbenzene; 23b). After treatment of (E)-22b (123 mg, 0.40 mmol) for 1.5 h at 2008, distillation

from a microflask at 1408/10^ꢁ3Torr yielded 23b (114 mg, 93%), which crystallized from EtOH in colorless

prisms. M.p. 59 – 618. UV (decalin): l_max261 (log e 2.91), 265 (2.92). IR (KBr): 1441m, 1489s, 1507s,

1

1592m (arom. ring vibr.); 1617w (CH¼CH, str.). H-NMR (400 MHz): 2.16 (s, Me); 2.79, 3.61 (2 dq,

H_aꢁC(5), H_bꢁC(5), ²J(a,b) ¼ 16.2; each of the two H-atoms shows similar couplings for ³J(5,1), ⁴J(5,2),

⁴J(5,3) ¼ 1.5 – 1.8, resulting in pseudo-q); 4.75 (s, broadened, HꢁC(3)); 5.97, 5.99 (2m, ³J(1,2) ¼ 5.9,

HꢁC(1), HꢁC(2)); 6.70 – 7.40 (7m, 14 arom. H). ¹³C-NMR (100 MHz, DEPT): 21.3 (Me); 46.7 (C(5));

59.7 (C(3)); 51.3 (C(4)); 125.7, 126.0, 128.3, 136.0 (2 p-CH of Ph₂, C(1), C(2)); 127.4, 128.15, 128.65,

129.58, 129.60, 129.76 (double intensity, 4 o- and m-CH of Ph₂, 2 o- and m-CH of p-tolyl); 135.8, 138.1,

145.9, 151.7 (4 arom. C_q). Anal. calc. for C₂₄H₂₂(310.42): C 92.86, H 7.14; found: C 92.84, H 6.99.

3.6.3. 3-(3-Chlorophenyl)-4,4-diphenylcyclopent-1-ene (¼1-Chloro-3-(5,5-diphenylcyclopent-2-en-1-

yl)benzene; 23c). Thermolysis (2008, 1.5 h) of (E)-22c (212 mg, 0.64 mmol) and distillation (1608/10^ꢁ3

Helvetica Chimica Acta – Vol. 94 (2011)

1385

Torr) furnished 23c (194 mg, 92%). Pale-yellow oil. UV (decalin): l_max261 (log e 3.40). ¹H-NMR

(60 MHz): 2.69, 3.60 (AB spectrum, br., J_gem¼ 16, H_aꢁC(5), H_bꢁC(5)); 4.73 (br. s, HꢁC(3)); 5.98 (br. s,

HꢁC(1), HꢁC(2)). Anal. calc. for C₂₃H₁₉Cl (330.84): C 83.49, H 5.79; found: C 83.86, H 5.76.

3.6.4. 3-(4-Methoxyphenyl)-4,4-diphenylcyclopent-1-ene (¼1-(5,5-Diphenylcyclopent-2-en-1-yl)-4-

methoxybenzene; 23d). Compound (E)-22d was subjected to the usual thermolysis and distillation

(1508/10^ꢁ3Torr) and afforded 23d (90%). M.p. 99 – 1018 (EtOH). UV (decalin): l_max279 (log e 3.23). IR

(KBr): 1020vs (CꢁOꢁC, sym. str.); 836s, 863s (CꢁOꢁC, asym. str.); 1440s, 1506vs, 1580m, 1610s (arom.

1

2

breath. modes). H-NMR (400 MHz): 2.77, 3.59 (2 dq, J(5a,5b) ¼ 16.2, three further small couplings,

H_aꢁC(5), H_bꢁC(5)); 3.68 (s, MeO); 4.74 (pseudo-s, HꢁC(3)); 5.96, 5.99 (2m, 14 lines visible, ³J(1,2) ¼ 5.9,

HꢁC(1), HꢁC(2)); 6.55 (dt, ³J(2’,3’) ¼ 6.7, 2 HꢁC(3’/5’) of MeOC₆H₄); 6.75 (dt, ³J(2’,3’) ¼ 6.7, 2 HꢁC(2’/

6’) of MeOC₆H₄); 6.85 – 7.40 (5m, 10 arom. H). ¹³C-NMR (100 MHz, DEPT): 46.6 (C(5)); 55.5 (MeO);

59.2 (C(3)); 61.4 (C(4)); 113.4 (C(3’/5’) of MeOC₆H₄); 133.3, 145.9, 151.6, 158.2 (4 arom. C_q). Anal. calc.

for C₂₄H₂₂O (326.42): C 88.30, H 6.79; found: C 88.57, H 6.86.

3.6.5. 3-(4-Nitrophenyl)-4,4-diphenylcyclopent-1-ene (¼1-(5,5-Diphenylcyclopent-2-en-1-yl)-4-nitro-

benzene; 23e). The thermolyzed (1.5 h, 2008) product was distilled at 1908/10^ꢁ3Torr; 92%, m.p. 80 – 858;

yellow-brown prisms from EtOH, m.p. 84 – 868. UV (decalin): l_max274 (log e 3.76). IR (ATR): 833m,

849m (p-disubst. C₆H₄, oop); 1335s, 1509vs (NO₂, sym. and asym. str.); 1442m, 1492s, 1595m (arom. ring

vibr). ¹H-NMR (400 MHz): 2.81, 3.63 (2dq, ²J(5a,5b) ¼ 16.5, H_aꢁC(5), H_bꢁC(5)); 4.85 (pseudo-s,

HꢁC(3)); 5.97, 6.10 (2dq, ³J(1,2) ¼ 5.9, HꢁC(1), HꢁC(2)); 6.80 – 7.36 (12 arom. CH); 7.85 (d, further split,

HꢁC(3’,5’) of NO₂C₆H₄). ¹³C-NMR (100 MHz, DEPT): 46.8 (C(5)); 59.8 (C(3)); 61.9 (C(4)); 150.8,

149.5, 145.0 (3 arom. C_q). Anal. calc. for C₂₃H₁₉NO₂(341.39): C 80.91, H 5.61, N 4.10; found: C 81.20, H

5.65, N 4.15.

3.6.6. 4,4-Diphenyl-3-{4-[(E)-styryl]phenyl}cyclopentene (¼1-(5,5-Diphenylcyclopent-2-en-1-yl)-4-

[(E)-2-phenylethenyl]benzene; 23f). Thermolysis of 22f at 2008 for 1.5 h gave 23f (96%). M.p. 115 – 1178

(EtOH). UV (decalin): l_max302 (log e 4.20), 316 (4.23), 330 (sh). IR (KBr): 695vs, 748s, 756m (Ph, oop);

820 (p-disubst. C₆H₄, oop); 960s (CH¼CH, oop). ¹H-NMR (400 MHz): 2.80, 3.63 (2dq, ²J(5a,5b) ¼ 16.0,

3

H_aꢁC(5), H_bꢁC(5)); 4.78 (pseudo-s, HꢁC(3)); 5.98, 6.03 (2dq, J(1,2) ¼ 5.8, HꢁC(1), HꢁC(2)); 6.80 –

7.50 (m, 19 arom. H, 2 olefin. H). ¹³C-NMR (100 MHz, DEPT): 46.8 (C(5)); 59.9 (C(3)); 61.5 (C(4));

8 double-intensity signals for pairs of o- and m-CH; 5 arom. C_q. Anal. calc. for C₃₁H₂₆(398.52): C 93.42, H

6.58; found: C 93.49, H 6.47.

3.6.7. 3-(Naphthalen-1-yl)-4,4-diphenylcyclopentene (¼1-(5,5-Diphenylcyclopent-2-en-1-yl)naphtha-

lene; 23g). The usual rearrangement of (E)-22g yielded 23g (94%). Glistening leaflets. M.p. 135 – 1368

(EtOH). UV (decalin): l_max2.87 (log e 3.45), sh at 276 and 299 nm. IR (KBr): 1492s, 1445s, 1490s, 1595m

(arom. ring vibr.). ¹H-NMR (400 MHz): 2.83, 3.78 (2dq, ²J(5a,5b) ¼ 17.1, further split, H_aꢁC(5),

H_bꢁC(5)); 5.62 (s, broadened, HꢁC(3)); 6.04, 6.10 (2m, ³J(1,2) ¼ 5.8, HꢁC(1), HꢁC(2)); 6.66 – 8.18 (8m,

17 arom. H). ¹³C-NMR (100 MHz, DEPT): 46.8 (C(5)); 54.6 (C(3)); 61.2 (C(4)); 132.6, 134.0, 137.8,

145.0, 152.3 (5 arom. C_q). Anal. calc. for C₂₇H₂₂(346.45): C 93.60, H 6.40; found: C 93.76, H 6.32.

3.6.8. 3-(Anthracen-9-yl)-4,4-diphenylcyclopentene (¼ 9-(5,5-Diphenylcyclopent-2-en-1-yl)anthra-

cene; 23h). Compound (E)-22h (112 mg, 0.28 mmol) in dioxan (2 ml) was sealed in a thick-walled

ampoule and heated 14 h at 1608; from EtOH pale-yellow leaflets (103 mg, 93%) of 23h. M.p. 135 – 1378

(recryst. EtOH). UV (decalin): l_max261 (log e 4.60), 352 (3.34), 370 (3.60), 391 (3.57). IR (KBr): 1621w

1

2

((E)-CH¼CH, str.). H-NMR (400 MHz): 3.09 (br. d, J(5a,5b) ¼ 17.1, H_aꢁC(5)); 4.25 (dd, J(5a,5b) ¼

17.6, J ¼ 3.4, H_bꢁC(5)); 6.14 (s, HꢁC(1/2)); 6.41 (br. s, HꢁC(3)); 6.55 (t, 2 H), 6.64 (t, 1 H), and 6.83

(t, 2 H) are tentatively assigned to m-H, p-H, and o-H of cis-4-Ph, shifted upfield by the ring current of

one side-ring of anthracen-9-yl; the ms at 7.05 – 7.43 (9 H) are attributed to trans-4-Ph and the anthracen-

9-yl H-atoms HꢁC(1’) to HꢁC(4’), the latter being shielded by the neighboring cis-4-Ph; five ¹H signals

come from the ꢄundisturbedꢂ second side-ring of anthracen-9-yl: 7.74 (d, ³J ¼ 8.00, HꢁC(5’)), 8.05 (d, ³J ¼

9.08, HꢁC(8’)), 7.94, 8.31 (2m, HꢁC(6’/7’), 8.21 (s, HꢁC(10’)). ¹³C-NMR (100 MHz, DEPT): 49.3 (C(5));

59.8 (C(3)); 60.0 (C(4)); 4 double-intensity signals of arom. CH are consistent with o- and m-CH of 2 Ph.

Anal. calc. for C₃₁H₂₄(396.50): C 93.90, H 6.10; found: C 93.84, H 6.10.

3.6.9. Rate Measurements of Ring Enlargement. The ampoule technique was the same as described

for the racemization, but the spectrophotometry required smaller concentrations. In a typical run, 22a

(1.30 mg, 4.39 · 10^ꢁ3mmol) was dissolved in decalin (25 ml). The soln. (0.176 mm) was clarified by

1386

Helvetica Chimica Acta – Vol. 94 (2011)

filtering and filled in 2-ml portions into 12 ampoules which were sealed under N₂. After defined reaction

times in the thermostat at 159.38, the ampoules were cooled, opened, and the absorbance was measured

in a 0.5-cm quartz cuvette, using the UV spectrometer RPQ 20C (Zeiss). From A_o¼ 1.02, the absorbance

A_tat 269 nm decreased to 0.201 (80% reaction) within 360 min. A₁¼ 0.076 was observed after 25 h, in

10.7 half-times. The evaluation of ten measurements of A_tby Eqn. 2 gave k_isom¼ 9.36 · 10^ꢁ5s^ꢁ1with R ¼

0.999.

The final A value deviated by a factor of 2.9 from A₁¼ 0.026, precalculated from the absorbance

coefficients of 22,500 (for 22a) and 450 (for 23a) at 269 nm. This deviation amounts to factor 2.0 – 3.3 in

the eight measurements listed in Table 7, and is caused by a small percentage of side reactions, which are

not detectable in the ¹H-NMR spectrum. The neglect of the side reaction may bring a small systematic

error into the numerical results, due to the seeming acceleration of the rate by parallel reactions. The k_isom

values of double runs agree within 1.5%.

The activation parameters (Table 7) were calculated from measurements at 159.3 – 184.08 with R ¼

0.995 – 0.999. With an estimated uncertainty of ꢀ 5% in k_isomand ꢀ 0.38 in the temp., the tolerances given

in Table 7 resulted.

3.6.10. ¹H-NMR-Spectroscopic Measurement of Concentration. The parent compound (E)-22a

(100 mg, 0.34 mmol) was dissolved in decalin (0.5 ml), sealed under N₂in a NMR tube, and brought into

Table 9. X-Ray Crystallographic Data of Compounds 22a and 23a

22a (Jomu 002)

23a (Jomu 001)

Empirical formula

Formula weight

Temp. [K]

C₂₃H₂₀

296.39

100(2)

C₂₃H₂₀

296.39

100(2)

Wavelength [ꢅ]

Crystal system, space group

Unit cell dimensions:

a [ꢅ]

b [ꢅ]

c [ꢅ]

a [8]

b [8]

g [8]

V [ꢅ³], Z

Calc. density [mg/mm³]

Absorption coefficient [mm^ꢁ1

F(000)

0.71073

Monoclinic, P2₁

0.71073

triclinic, P1

9.8543(2)

7.7983(2)

10.8859(2)

11.4326(4)

11.7904(4)

12.8679(4)

81.153(2)

89.214(2)

71.542(2)

1624.59(9), 4

1.212

94.2810(10)

834.21(4), 2

1.180

0.066

]

0.068

632

316

Crystal size [mm]

V Range [8]

Index ranges

0.30 ꢃ 0.29 ꢃ 0.28

2.07 – 30.05

ꢁ 13 ꢄ h ꢄ 13

ꢁ 10 ꢄ k ꢄ 10

ꢁ 15 ꢄ l ꢄ 15

33412

0.30 ꢃ 0.30 ꢃ 0.20

1.60 – 30.07

ꢁ 16 ꢄ h ꢄ 16

ꢁ 16 ꢄ k ꢄ 16

ꢁ 18 ꢄ l ꢄ 18

105925

Reflections collected

Reflections unique

2608

9501

R(int)

0.0459

0.0550

Completeness to 2V¼ 30.05

Data/restraints/parameters

Goodness-of-fit on F²

Final R [I > 2s(I)], wR2

Largest diff. peak/hole [e ꢅ^ꢁ3

CCDC Deposition No.

100.0%

2608/1/208

1.037

0.0351, 0.0876

0.239/ ꢁ 0.190

785107

99.6%

9501/0/415

1.027

0.0446, 0.1088

0.362/ ꢁ 0.219

785106

]

Helvetica Chimica Acta – Vol. 94 (2011)

1387

a preheated bath of polyethyleneglycol (177.58). In suitable intervals, the 100-MHz spectra were taken,

and the signals at 5.46 ppm for the HꢁC(1’) of 22a and at 6.00 for the two vinyl H-atoms of 23a were

isolated from the printed spectra and weighed (F₂₂and F₂₃). The percentage of F₂₂in F₂₂þ F₂₃was

introduced in the first-order Eqn. 1; 12 measurements up to 77% rearrangement furnished k_isom¼ 3.85 ·

10^ꢁ4s^ꢁ1. That is 11% less than 4.32 · 10^ꢁ4s^ꢁ1, measured by spectrophotometry, but the very different

concentrations of 22a in the decalin soln. (6.2 · 10^ꢁ5% vs. ca. 20%) changed the medium.

3.6.11. Solvent Dependence (Table 4). It is noteworthy that no interaction with alcohols took place

during the isomerization. In a prep. experiment, 22a (117 mg) in ethane-1,2-diol (100 ml) was heated for

10 h at 1708. Workup with H₂O/Et₂O gave 23a (108 mg, 92%). For the kinetic runs in low-boiling solvents,

thick-walled ampoules were used. The gap between the exper. and the precalculated absorbance A₁

appears to depend on the purity of the solvents; in benzene, PhCl, dioxan etc., there was hardly a

difference noticeable.

4. Crystallographic Structure Determination. X-Ray diffraction measurements were performed on a

Bruker X8 APEX II CCD-diffractometer. Single crystals of compounds 22a and 23a were mounted on

glass fibers, coated with Parathone N oil and positioned at 40 mm from the detector. 2271 and 3470

frames were measured, each for 20 s over 18 scan width for 22a and 23a, resp. The data were processed

using SAINT software [58]. Crystal data, data collection parameters, and structure refinement details for

22a and 23a are given in Table 9. The structures were solved by direct methods and refined by full-matrix

least-squares techniques. Non-H-atoms were refined with anisotropic displacement parameters, while the

H-atoms were placed at geometrically calculated positions and refined as riding atoms in the subsequent

least-squares model refinements. The isotropic thermal parameters were estimated to be 1.2 times the

values of the equivalent isotropic thermal parameters of the atoms to which H-atoms were bonded. The

following computer programs and computer were used: structure solution, SHELXS-97 [59]; refinement,

SHELXL-97 [60]; molecular diagrams, ORTEP [61]; computer: Pentium IV.

REFERENCES

[1] E. Vogel, Angew. Chem. 1960, 72, 4.

[2] C. G. Overberger, A. E. Borchert, J. Am. Chem. Soc. 1960, 82, 4896.

[3] M. C. Flowers, H. M. Frey, J. Chem. Soc. 1961, 3547.

[4] D. K. Lewis, D. S. Charney, B. L. Kalva, A.-M. Plate, M. H. Woodard, S. J. Cianciosi, J. E. Baldwin,

J. Phys. Chem. A 1997, 101, 4097.

[5] E. W. Schlag, B. S. Rabinovitch, J. Am. Chem. Soc. 1960, 82, 5996.

[6] H. G. Korth, H. Trill, R. Sustmann, J. Am. Chem. Soc. 1981, 103, 4483.

[7] R. B. Woodward, R. Hoffmann, Angew. Chem. 1969, 81, 797; Angew. Chem., Int. Ed. 1969, 8, 781.

[8] M. R. Willcott, V. H. Cargle, J. Am. Chem. Soc. 1967, 89, 723.

[9] W. v. E. Doering, K. Sachdev, J. Am. Chem. Soc. 1974, 96, 1168; W. v. E. Doering, K. Sachdev, J. Am.

Chem. Soc. 1975, 97, 5512.

[10] J. E. Baldwin, Chem. Rev. 2003, 103, 1197.

[11] G. D. Andrews, J. E. Baldwin, J. Am. Chem. Soc. 1976, 98, 6705.

[12] J. E. Baldwin, K. A. Villarica, D. I. Freedberg, F. A. L. Anet, J. Am. Chem. Soc. 1994, 116, 10845.

[13] L. A. Asuncion, J. E. Baldwin, J. Am. Chem. Soc. 1995, 117, 10672.

[14] K. N. Houk, Y. Li, J. D. Evanseck, Angew. Chem. 1992, 104, 711; Angew. Chem., Int. Ed. 1992, 31,

682.

[15] K. N. Houk, M. Nendel, O. Wiest, J. W. Storer, J. Am. Chem. Soc. 1997, 119, 10545.

[16] E. R. Davidson, J. J. Gajewski, J. Am. Chem. Soc. 1997, 119, 10543.

[17] M. Nendel, D. Sperling, O. Wiest, K. N. Houk, J. Org. Chem. 2000, 65, 3259.

[18] A. Ohta, K. Dahl, R. Raab, J. Geittner, R. Huisgen, Helv. Chim. Acta 2008, 91, 783.

[19] W. Sicking, R. Sustmann, R. Huisgen, J. Mulzer, Helv. Chim. Acta 2011, 94, doi: 10.1002/

hlca.201100142.

[20] H. M. Walborsky, F. M. Hornyak, J. Am. Chem. Soc. 1955, 77, 6026.

[21] H. C. Brown, A. Tsukamoto, J. Am. Chem. Soc. 1961, 83, 4549.

1388

Helvetica Chimica Acta – Vol. 94 (2011)

[22] G. A. R. Kon, J. Chem. Soc. 1948, 224.

[23] J. Mulzer, R. Huisgen, R. Sustmann, in preparation.

[24] E. Pretsch, T. Clerc, J. Seibl, W. Simon, ꢄTabellen zur Strukturaufklꢁrung organischer Verbindungen

mit spektroskopischen Methodenꢂ, Springer-Verlag, Berlin, Heidelberg, New York, 1976.

[25] H. M. Walborsky, private communication to R. H.

[26] H. M. Walborsky, L. Barash, A. E. Young, F. J. Impastato, J. Am. Chem. Soc. 1961, 83, 2517.

[27] C. Reichardt, Chem. Rev. 1994, 94, 2319.

[28] B. F. Yates, W. J. Bouma, L. Radom, Tetrahedron 1986, 42, 6225.

[29] Computer program DavinX: D. S. Stephenson, ꢄEncyclopedia of Nuclear Magnetic Resonanceꢂ,

Eds. T. M. Grant, R. K. Harries, John Wiley & Sons, New York, 1996, pp. 816 – 821.

[30] A. de Meijere, W. Lꢀttke, Tetrahedron 1969, 25, 2047.

[31] B. H. Northrop, K. N. Houk, J. Org. Chem. 2005, 71, 3.

[32] G. R. de Mare, J. S. Martin, J. Am. Chem. Soc. 1966, 88, 5033; H. Gꢀnther, D. Wendisch, Angew.

Chem. 1966, 78, 266; Angew. Chem., Int. Ed. 1966, 5, 251.

[33] M. L. Coote, C. Y. Lin, H. Zipse, ꢄThe Stability of Carbon-Centered Radicalsꢂ, in ꢄCarbon-Centered

Free Radicals and Radical Cationsꢂ, Ed. M. D. E. Forbes, John Wiley & Sons, 2010, Chapt. 5.

[34] D. H. Hey, Adv. Free Radical Chem. 1967, 2, 47.

[35] R. Huisgen, F. Jakob, R. Grashey, Chem. Ber. 1959, 92, 2206.

[36] R. Ito, T. Migita, N. Morikawa, O. Simamura, Bull. Soc. Chem. Jpn. 1963, 36, 992.

[37] S. C. Dickermann, G. B. Vermont, J. Am. Chem. Soc. 1962, 84, 4150.

[38] J. M. Simpson, H. G. Richey, Tetrahedron Lett. 1973, 27, 2545.

[39] H. G. Richey, D. W. Shull, Tetrahedron Lett. 1976, 30, 575.

[40] M. Buchert, H.-U. Reissig, Liebigs Ann. 1996, 2007.

[41] P. Caramella, R. Huisgen, B. Schmolke, J. Am. Chem. Soc. 1974, 96, 2997; P. Caramella, R. Huisgen,

B. Schmolke, J. Am. Chem. Soc. 1974, 96, 2999.

[42] J. Chateauneuf, J. Lusztyk, K. U. Ingold, J. Org. Chem. 1988, 53, 1629.

[43] C. Rꢀchardt, H.-D. Beckhaus, Angew. Chem. 1980, 92, 417; Angew. Chem., Int. Ed. 1980, 19, 429; C.

Rꢀchardt, H.-D. Beckhaus, Angew. Chem. 1985, 97, 531; Angew. Chem., Int. Ed. 1985, 24, 529.

[44] D. Griller, K. U. Ingold, Acc. Chem. Res. 1976, 9, 13.

[45] C. Rꢀchardt, H.-D. Beckhaus, Top. Curr. Chem. 1986, 130, 1.

[46] M. Gomberg, J. Am. Chem. Soc. 1900, 22, 757; M. Gomberg, Ber. Dtsch. Chem. Ges. 1900, 33, 3150.

[47] H. Lankamp, T. W. Nauta, C. MacLean, Tetrahedron Lett. 1968, 249.

[48] K. Ziegler, P. Orth, K. Weber, Liebigs Ann. Chem. 1933, 504, 131.

´

[49] S. S. Kristjansdottir, A. E. Moody, J. R. Norton, Int. J. Chem. Kinet. 1992, 24, 895.

[50] K. Ziegler, L. Ewald, Liebigs Ann. Chem. 1929, 473, 163.

[51] T. Colle, P. S. Glaspie, E. S. Lewis, J. Org. Chem. 1978, 43, 2722.

[52] M. Stein, W. Winter, A. Riecker, Angew. Chem. 1978, 90, 737; Angew. Chem., Int. Ed. 1978, 17, 692.

[53] E. W. Yankee, F. D. Badea, N. E. Howe, D. J. Cram, J. Am. Chem. Soc. 1973, 95, 4210.

[54] N. E. E. Howe, E. W. Yankee, D. J. Cram, J. Am. Chem. Soc. 1973, 95, 4230.

[55] R. Hoffmann, J. Am. Chem. Soc. 1968, 90, 1475.

[56] H. Staudinger, E. Anthes, F. Pfenninger, Chem. Ber. 1916, 49, 1928; J. B. Miller, J. Org. Chem. 1959,

24, 560.

[57] K. Friedrich, A. Henning, Chem. Ber. 1959, 92, 2756.

[58] SAINT-Plus (Version 7.06a) and APEX2, Bruker-Nonius AXS Inc., 2004, Madison, Wisconsin.

[59] G. M. Sheldrick, SHELXS-97, Program for Crystal Structure Solution, University Gçttingen,

Gçttingen, 1997.

[60] G. M. Sheldrick, SHELXL-97, Program for Crystal Structure Refinement, University Gçttingen,

Gçttingen, 1997.

[61] G. K. Johnson, Report ORNL-5138, Oak Ridge National Laboratory, Oak Ridge, TN, 1976.

Received April 5, 2011

Article Doi

1-[(E)-2-arylethenyl]-2,2-diphenylcyclopropanes: Kinetics and mechanism of rearrangement to cyclopentenes

DOI: 10.1002/hlca.201100135

Source and publish data:

Authors:

Article abstract of DOI:10.1002/hlca.201100135

Full text of DOI:10.1002/hlca.201100135

Products guided by the article

R&D Labs maybe for 1333307-70-2

Relevant to this article

Hot Product