New results in homoheptalene chemistry

Helvetica Chimica Acta ± Vol. 84 (2001)

1017

New Results in Homoheptalene Chemistry

by Georg Rüedi¹) and Hans-Jürgen Hansen*

Organisch-chemisches Institut der Universität Zürich, Winterthurerstrasse 190, CH-8057 Zürich

The thermal reaction of homoazulene ( bicyclo[5.3.1]undeca-1,3,5,7,9-pentaene; 2) with dimethyl

acetylenedicarboxylate (ADM) in 1,2-dichloroethane (ClCH₂CH₂Cl) results, in contrast to an earlier report

[5], in formation of not only dimethyl homoheptalene-4,5-dicarboxylate ( bicyclo[5.5.1]trideca-1,3,5,7,9,11-

hexaene-4,5-dicarboxylate; 3), but also of a 4 :1 mixture of 3 and dimethyl homoheptalene-2,3-dicarboxylate

(13) in almost quantitative yield (Schemes 1 and 3). The structures of both homoheptalenes have been

corroborated by X-ray crystal-structure analysis (Fig. 5). The double-bond-shifted (DBS) isomers 3' and 13' of

3 and 13, respectively, could not be detected in their ¹H-NMR spectra (600 MHz threshold of detection ꢀ0.5%),

in agreement with the AM1-calculated DH^o_fvalues of the four isomeric homoheptalene-dicarboxylates (cf.

Table 4). Vilsmeyer formylation of homoazulene (2) gave homoazulene-8-carbaldehyde (14) in a yield of 67%,

which, on treatment with benzylidene-(triphenyl)-l⁵-phosphane, gave, in almost quantitative yield, a 1.6 :1

mixture of (Z)- and (E)-8-styrylhomoazulene ((Z)-15 and (E)-15, resp.). Thermal reaction of the latter mixture

with ADM in 1,2-dichloroethane led, in a yield of 42%, to a 5 :1 mixture of dimethyl (Z)- and (E)-2-

styrylhomoheptalene-4,5-dicarboxylate ((Z)-15 and (E)-16, resp.). Both isomers were separated by column

chromatography on silica gel. Again, the DBS isomers of (Z)-16 and (E)-16, i.e., (Z)-16' and (E)-16', could not

1

be detected in the H-NMR spectra (600 MHz) of pure (Z)-16 and (E)-16.

1. Introduction. ± The class of homoheptalenes ( bicyclo[5.5.1]trideca-1,3,5,7,9,11-

hexaenes) is still sparsely populated, although the parent structure was synthesized more

than 25 years ago by Vogel et al. [1]. Line-shape analysis of the ¹³C-NMR signals of

homoheptalene (1a) in the temperature range of 153 ± 233 K gave an activation energy

1

(E_a) of 5 kcal ´ mol for the double-bond-shift (DBS) process in this [12]annulene,

1

which is slightly higher than that for heptalene itself (3.5 kcal ´ mol ), which was

derived by the same procedure [2]. Two X-ray crystal-structure analyses of the

homoheptalene derivatives 1b [3] and 1c [4] have been performed (see below) that

show the p-core to be composed of two cycloocta-2,4,6-triene substructures with the

ꢀhomo-C(sp³)-atomꢁ and its two attached C(sp²)-atoms in common and with a c₂axis

passing through the bridging C-atom (see the AM1-calculated structure of 1a in Fig. 1,

which is in good agreement with the X-ray crystal structures of 1b and 1c (Sect. 2.2).

Later, Scott and Kirms applied Hafnerꢁs synthesis of dimethyl heptalene-4,5-dicarbox-

ylates by thermal reaction of azulenes and dimethyl acetylene-dicarboxylate (ADM)

(cf. [5]) to homoazulene (2), for which Scott et al. had developed a new and efficient

synthesis [6]. They reported that 2 reacts smoothly with ADM in boiling 1,2-

dichloroethane (ClCH₂CH₂Cl) to yield >90% dimethyl homoheptalene-4,5-dicarbox-

ylate (3)²) (Scheme 1) [5]. Apart from the specific case of benzo[b]homoheptalene [7],

1

)

Taken in part from the diploma thesis of G. R., University of Zurich, 2000.

The locants of homoheptalene (1a) are based on the systematic name: tricyclo[5.5.1]trideca-1,3,5,7,9,11-

hexaene.

2

1018

Helvetica Chimica Acta ± Vol. 84 (2001)

homoheptalene 3 seems, so far, the only known unsymmetrically substituted

homoheptalene that, in principle, could exist in two double-bond-shifted (DBS) forms.

1

However, in the H-NMR spectrum of 3, Scott and Kirms could recognize only the

signals that were in agreement with the structure of a 4,5-dicarboxylate.

Fig. 1. AM1-Calculated structure of homoheptalene (1a)

F

Br

1b

1c

Scheme 1

ADM

E

82°/ 30 h

E

ClCH₂CH₂Cl

2

3 (> 90 %)

In the context of our investigations on the DBS isomers of heptalenes as two

switchable p-states [8], we also became interested in DBS isomers of unsymmetrically

substituted homoheptalenes and decided to repeat the Scott and Kirmsꢁs synthesis of 3

because we had acquired some skill in the derivatization of heptalene-4,5-dicarbox-

ylates (cf. [9]). In the following, we give a first report of our experiments with

homoazulene (2) and homoheptalene-dicarboxylates derived therefrom.

2. Results and Discussion. ± 2.1. Homoazulene (2). The main steps of the synthesis

of homoazulene (2) of Scott et al. are displayed in Scheme 2 [6]. The crucial step of the

synthesis is based on a catalyzed, intramolecular Buchner reaction of the diazo ketone

3

4, which is quantitatively available from dihydrocinnamoyl chloride and CH₂N₂). For

the formation of 3,4-dihydro-2H-azulen-1-one (5), we followed an improved

procedure of Scott and Sumpter [10], who obtained 5 as a slightly green oil in 77%

yield in boiling CH₂Cl₂in the presence of catalytic amounts of [Rh₂(OAc)₄], ensued by

3

)

We performed all reactions on a larger scale than reported by Scott et al. in their synthesis of 2. This

required, at several steps, modifications of the experimental techniques described in the Exper. Part.

Helvetica Chimica Acta ± Vol. 84 (2001)

1019

Scheme 2

O

X

a)

b)

CHN₂

4

5

X = O : 6

c)

X = NNHTs : 7

OAc

d)

g)

e)

f)

OAc

9

10

8

i)

OR

R = H : 11

2

h)

R = Ms : 12

a) Rh(OAc)₄, CH₂Cl₂; 61 ± 69%. b) Me₃SO I , NaH, DMSO, 758; 56 ± 62%. c) TsNHNH₂, THF; 87%. d) MeLi,

Et₂O; 88 ± 91%. e) Pb(OAc)₄, benzene, AcOH; 28%. f) Pd(OAc)₂, Ph₃P, Na₂CO₃, heptane, 858; 68 ± 96%. g)

MeLi, Et₂O; 90%. h) MsCl, Et₃N, CH₂Cl₂; 97%. i) DBN, Et₂O; 50 ± 53%

base-catalyzed DBS. We purified 5 by column chromatography on silica gel and

crystallized the material from hexane/t-BuOMe 20 :1. In this way, we obtained pure 5 in

yields of 61 ± 69% in colorless crystals (m.p. 298). The cyclopropanation of 5 with

trimethylsulfoxonium iodide in the presence of NaH in dimethyl sulfoxide (DMSO),

according to a procedure originally described by Corey and Chaykowsky [11], led to the

tricyclic ketone 6 in yields of 56 ± 62%, which are lower than the yield (76%) reported

by Scott et al.⁴). The next important step is the Shapiro reaction of 7, the tosyl

hydrazone of 6, with 2 mol-equiv. of MeLi in Et₂O. We obtained the tricyclic triene 8 in

yields of 88 ± 91%, which are almost the same as the reported yield of 92%.

Nevertheless, we performed our experiments in 10-fold larger runs. The next step,

namely the oxidative cleavage of the three-membered ring of 8 by Pb(AcO)₄in

benzene in the presence of AcOH to the bicyclic diacetate 9, was found to be the real

obstacle in the synthesis of 2. Scott et al. reported a yield of 73% for this step ± a yield

that we could neither reach nor approximate, also in applying procedures of other

authors (cf. [12]). The results of a number of our experiments are collected in Table 1.

4

)

A control experiment performed exactly under the conditions of Scott et al. led also to a yield of 76% of 6.

However, we worked in larger-scale runs with five-fold higher concentrations in DMSO to avoid the use of

too-huge quantities of DMSO. Moreover, we reduced the amount of trimethylsulfoxonium iodide applied

from 3 to 1.3 mol-equiv. per mol-equiv. of 5.

1020

Helvetica Chimica Acta ± Vol. 84 (2001)

Table 1. Acetoxylation of Tricyclo[5.3.1.0^1,7]undeca-2,4,9-triene (8) with Pb(OAc)₄

Pb(OAc)₄

Solvent

Time [h]

Temp. [8]

Yield [%]

1.0 equiv.

3.0 equiv.

1.5 equiv.

2.0 equiv.

Benzene/AcOH 4 : 1

Benzene

AcOH

Benzene/AcOH 5 : 2

4

7

5

6

5

10

9

25

60

25

100

25

45

15

17

15

18

20

0

23

28

The best yield of 9 obtained by us amounted to 28%. Nevertheless, we obtained 9 after

chromatography and crystallization from hexane/t-BuOMe as colorless crystals,

suitable for X-ray crystal-structure analysis (Fig. 2), which unambiguously confirmed

its structure. The elimination of 1 equiv. of AcOH from 9 with catalytic amounts of

Pd(OAc)₂in the presence of Ph₃P and Na₂CO₃in heptane at 858 caused no specific

problems, and acetate 10 was obtained in yields of 68 ± 96%, in agreement with the

reported yields of 71 ± 95%. Acetate 10 was obtained as a bright yellow, quite unstable

oil, as were the subsequent intermediates 11 and 12, so that the ensuing steps had to be

performed immediately one after the other. In this way, we achieved yields of ꢀ 90%

for each step, in agreement with the yields reported by Scott et al. [6]. The latter applied

for the last step, the elimination of MsOH from methanesulfonate 12, t-BuOK in t-

BuOH and finally obtained 2 in a yield of 30%. We found that the elimination reaction

with 12 can be performed much more smoothly with a seven-fold excess of 1,5-

diazabicyclo[4.3.0]non-5-ene (DBN) in Et₂O at ambient temperature to give pure 2 in

yields of 50 ± 53%. Fig. 3 shows the ¹H-NMR spectrum of 2 in (D₆)acetone at 600 MHz

with the assignment of all ¹H-NMR signals. It is noteworthy that H_aC(11) exhibits a

Fig. 2. Stereoscopic view of the crystal structure of bicyclo[5.3.1]undeca-2,4,9-triene-1,7-diyl diacetate (9)

Helvetica Chimica Acta ± Vol. 84 (2001)

1021

Fig. 3. ¹H-NMR Spectrum of homoazulene (2) (600 MHz; (D₆)acetone)

4

small W-like J coupling of 1.6 Hz with H C(8,10), which is not recognizable for

H_bC(11) and H C(2,6)⁵).

2.2. Formation and Structure of Homoheptalenes. We performed the reaction of 2

with 3-mol-equiv. of ADM in ClCH₂CH₂Cl at 828 as described by Scott and Kirms [5],

monitoring by GC/MS, and were quite excited to find two products in a ratio of 4 :1 in

the reaction mixture, each with the expected mass of m/z 284. Our first idea was that the

second product may represent the desired DBS isomer of 3, namely dimethyl

homoheptalene-3,4-dicarboxylate (3'). After 30 h, the mixture of the two products was

1

collected as a blood-red oil and in a total yield of 96%. The H-NMR spectrum

(CDCl₃) of the mixture indicated, indeed, the presence of two dimethyl homohepta-

lene-dicarboxylates in a ratio of 4 :1, whereby the s at 7.19 ppm, attributable to the

main isomer, was identical to the reported signal of H C(6) of 3. Repeated

chromatography of the mixture on a large amount of silica gel with hexane/t-BuOMe

4 :1 effected enrichment of both compounds, so that further purification was possible

1

by crystallization from hexane/t-BuOMe mixtures. The complete analyses of the H-

and ¹³C-NMR spectra of both compounds (cf. Fig. 4 and Table 2) revealed that the

second homoheptalene derivative, formed in minor amounts, was dimethyl homo-

heptalene-2,3-dicarboxylate (13; Scheme 3).

1

Scottꢁs assignment of the structure of 3 on the basis of its H-NMR spectrum at

1

100 MHz, which did not allow perfect resolution of all H signals, could be fully

Scheme 3

ADM

+

E

82°/ 30 h

E

DCE

96 %

(4 : 1)

2

3

13

5

)

The locants of 2 are based on its systematic name: bicyclo[5.3.1]undeca-1,3,5,7,9-pentaene.

1022

Helvetica Chimica Acta ± Vol. 84 (2001)

Fig. 4. ¹H-NMR Spectra (600 MHz; CDCl₃) of a) dimethyl homoheptalene-4,5-dicarboxylate and b) dimethyl

homoheptalene-2,3-dicarboxylate (3 and 13, resp.)

Table 2. ¹H-NMR Data (600 MHz, CDCl₃) of the Dimethyl Homoheptalene-dicarboxylates 3 and 13

Dicarb- H C(2) H C(3) H C(4)

oxylate

H

C(5) H C(6)

H

C(8)

H

C(9)

H

C(10)

H

C(11) H C(12)

H C(13)

3

5.92 (d, 6.99 (d, 3.71^a)

J 4.2) J 4.2)

3.68^a)

7.19 (s)

6.11

5.67 (dd,

5.60 (dd,

5.61 (dd, 6.09

5.07/4.97

(d, J 4.7) J 11.5, 6.6) J 11.5, 6.6) J 6.6)

(d, J 12.3) (J 11.3)

13

3.67^a)

3.69^a)

6.93

(d, J 6.4) J 12.8, (d, J 12.8) (d, J 5.2) J 12.2,

6.4) 5.2)

5.68 (dd, 5.99

5.99

5.86 (dd,

5.72 (dd,

J 12.2,

7.6)

6.00 (ddd, 7.71

5.41 (dd,

J 12.9, (d, J 12.9) J 11.1, 1.2)

7.6, 0.9)

4.55 (dd,

J 11.1,

0.8)

^a) MeOCO.

Helvetica Chimica Acta ± Vol. 84 (2001)

1023

confirmed. The appearance of the s of H C(6), already mentioned, as well as the d of

H

C(3) with ³J(3,2) 4.2 Hz in the low-field region at 7.19 and 6.99 ppm, respectively,

is typical for 3, and allows assignment of the positions of the CC bonds in this

[12]annulene. The chemical-shift difference between H_synC(13) and H_antiC(13) is

with 0.1 ppm quite small, in contrast to 13, where this difference is 0.86 ppm.

Nevertheless, the signal of H_synC(13), i.e., the H-atom positioned above the E_Me

-

substituted ring, appears in both isomers at lower field. Moreover, a closer inspection of

Fig. 4 reveals that the AB systems of CH₂(13) are located almost symmetrically around

5 ppm⁶), or, in other words, the magnitude of the low-field shift of H_synis almost

equivalent to that of the high-field shift of H_antiwith respect to the reference position of

5 ppm for both homoheptalenes. We believe that this observation supports the view

that the homotropylium-like structure 13a contributes to the description of the ground

state of 13, but much less so for 3 and the mesomeric structure 3a.

O

3

13

OMe

E

O

OMe

13a

3a

An observation that confused us at first was the appearance of a comparably low-

field d at 7.71 ppm, with ³J 12.9 Hz, in the ¹H-NMR spectrum of 13, until we learned

from NOESY and COSY experiments that it corresponded to H C(12). An X-ray

crystal-structure analysis of 13 as well as of 3 (cf. Fig. 5) showed clearly that, at least in

the crystal structure, H C(12) lies well in the deshielding area of the CO group of

E_MeC(2). Moreover, the crystal-structure conformation represents also the lowest-

energy conformation of the AM1-calculated structure of 13, and it must also be the

prevailing conformation, at least in CDCl₃solution, according to the observed chemical

shift of the signal of H C(12).

The most relevant X-ray crystal-structure data of 3 and 13 together with those of

the c₂-symmetric homoheptalenes 1b and 1c are compiled in Table 3. Added to these

data are the AM1-calculated structural parameters of the c₂-symmetric parent

compound 1a and, for the purpose of comparison, the X-ray crystal-structure data of

dimethyl heptalene-4,5-dicarboxylate (14) [13], the homoheptalene counterpart of

which is 13. The average length of the alternating C C and CC bonds of the 12e p-

perimeter of all structures are, as expected, very close together, and for 13, one finds

almost identical figures as for its heptalene analogue 14. The slight elongation of

d_av(CC) of 14 in comparison with d_av(CC) of 13 is with <1% within the limits of the

given precisions (see Table 3) and, therefore, of no significance. Nevertheless, it shows

the expected trend with respect to the distinctly smaller V_av(CC CC) and

V_av(C CC C) of 14 as compared with those of 13, which could be ascribed to a

6

)

The centers of the AB systems of 3 and 13 are at 5.02 and 4.98 ppm, respectively.

1024

Helvetica Chimica Acta ± Vol. 84 (2001)

Fig. 5. Stereoscopic views of the crystal structures of a) 3 and b) 13

stronger p-interaction in 14. The higher constraints of the homoheptalene structures

are also reflected in their average bond angles (V_av(CC C)), all of which are larger

than that of 14, specifically that of 1c (4%), which has almost the same

V_av(CC CC) value as for 14. On closer inspection of the structural parameters

of the homoheptalenes, one recognizes that substituents at the 12e p-perimeter have no

significant influence on the corresponding average data. Vicinal substituents as in 3 and

13 may alter the individual parameters of the substructures involved (cf., e.g., the

corresponding V values of 3 and 13, which are most susceptible to structural changes);

however, these deviations are compensated across the whole molecule due to the

structurally imposed constraints of internal motions. The situation changes when a

homoheptalene carries substituents at C(13), as is demonstrated by the structural data

of 1c with geminal F-substituents at C(13). Whereas d_av(CC) and d_av(C C) of the

perimeter are still comparable with those of the other homoheptalenes, already the

W_av(CC C) value is significantly larger than that of the homoheptalenes that were

subjected to X-ray crystal-structure analyses. Responsible for this observation is the

steric interaction of the F-substituents with the C(3)C(4) and C(8)C(10) bonds,

respectively, which reduces the V(C(3) C(4) C(5) C(6)) value by ꢀ198 as

compared with that of the other homoheptalenes. In other words, the F-substituents

at C(13) induce flattening of the two p-halves (C(2/8) C(3/9) C(4/10) C(5/

11) C(6/12)) of 1c, which, in turn, leads to enlargement of the bond angles involved,

as expressed as the W_av(CC C). That this factor reflects a general structural feature

and not the consequence of a specific structural behavior of the highly electronegative

F-atoms is demonstrated by an AM1 study of 13,13-X₂-substituted homoheptalenes,

the most relevant results of which are summarized in Table 4. The comparison of the

Helvetica Chimica Acta ± Vol. 84 (2001)

1025

Table 3. Comparison of Relevant Structural Data of Homoheptalenes and Dimethyl Heptalene-4,5-dicarboxylate

a

(14) )

b

c

d

e

f

Homoheptalene

1a )

1b )

1c )

3 )

13 )

14 )

Interatomic distances d [pm]

C(1) C(2)

C(1) C(12)

C(1) C(13)

C(2) C(3)

C(3) C(4)

C(4) C(5)

C(5) C(6)

C(6) C(7)

C(7) C(8)

C(7) C(13)

C(8) C(9)

C(9) C(10)

C(10) C(11)

C(11) C(12)

C(1) ´´´ C(7)

d_av(C C)

135.0

145.0

149.2

144.4

134.2

144.3

134.2

145.0

135.0

149.2

144.4

134.2

144.3

134.2

240.9

144.6

134.5

133.5(3)

145.0(3)

149.6(3)

145.9(4)

132.2(3)

146.6(3)

133.7(3)

145.0(3)

133.5(3)

149.6(3)

145.9(3)

132.2(3)

146.6(3)

133.7(3)

241.7

135.4(3)

145.8(3)

150.3(2)

144.9(3)

134.6(3)

144.7(3)

135.6(3)

145.4(3)

135.5(3)

150.3(3)

145.1(3)

134.2(3)

145.2(3)

134.7(3)

245.0

134.9(2)

145.5(2)

151.0(2)

145.3(2)

133.9(2)

149.8(2)

135.1(2)

145.0(2)

135.1(2)

150.4(2)

145.8(2)

133.8(2)

147.3(3)

134.7(2)

240.4

135.6(2)

145.2(3)

151.3(2)

149.4(2)

134.1(3)

146.3(3)

134.5(3)

146.5(3)

135.5(3)

150.6(3)

143.8(3)

135.1(3)

145.7(3)

135.3(3)

238.6

138.8

146.8

±

147.5

134.9

142.2

134.5

148.4

137.9

±

144.4

134.0

146.0

136.8

g

149.3 )

145.9

136.2

145.8

133.1

145.2

135.0

146.5

134.6

146.2

135.0

d_av(CC)

Bond angles W [8]

C(1) C(2) C(3)

C(1) C(13) C(7)

C(2) C(3) C(4)

C(3) C(4) C(5)

C(4) C(5) C(6)

C(5) C(6) C(7)

C(6) C(7) C(8)

C(7) C(8) C(9)

C(8) C(9) C(10)

C(9) C(10) C(11)

C(10) C(11) C(12)

C(11) C(12) C(1)

C(12) C(1) C(2)

W_av(CC C)

124.7

107.6

129.5

131.9

133.2

130.0

120.8

124.7

129.5

131.9

133.2

130.0

120.8

129.9

125.4

107.9

126.1

133.0

132.7

130.9

121.1

125.4

126.1

133.0

132.7

130.9

121.1

128.2

130.2(2)

109.2(2)

130.7(2)

132.8(2)

134.8(2)

132.4(2)

120.7(2)

131.1(2)

130.8(2)

132.9(2)

135.0(2)

132.5(2)

120.3(2)

130.4

123.6(2)

105.8(1)

128.3(1)

127.8(1)

131.7(1)

130.5(1)

120.1(1)

124.6(1)

128.4(2)

131.3(2)

134.3(2)

129.8(2)

121.3(1)

127.6

121.1(2)

104.5(1)

124.2(2)

129.0(2)

130.8(2)

130.6(2)

119.0(2)

125.8(2)

128.5(2)

131.7(2)

135.7(2)

129.6(2)

125.2(2)

127.6

125.4

±

125.8

126.1

127.0

124.2

120.5

129.8

124.0

126.3

128.3

124.3

123.5

125.4

Torsion angles V [8]

C(1) C(2) C(3) C(4)

C(2) C(3) C(4) C(5)

C(3) C(4) C(5) C(6)

C(4) C(5) C(6) C(7)

C(5) C(6) C(7) C(8)

C(6) C(7) C(8) C(9)

C(7) C(8) C(9) C(10)

C(8) C(9) C(10) C(11)

C(9) C(10) C(11) C(12)

C(10) C(11) C(12) C(1)

C(11) C(12) C(1) C(2)

C(12) C(1) C(2) C(3)

C(1) C(2) CO

49.5

5.2

37.7

1.7

163.5

179.0

49.5

5.2

37.7

1.7

163.5

179.0

±

47.7

8.2

41.3

2.2

169.6

175.7

47.7

8.2

41.3

2.2

169.6

175.7

±

40.1

2.3

22.0

4.1

155.4

174.5

40.1

2.3

22.0

4.1

155.4

174.5

±

48.8(2)

13.9(2)

48.9(2)

0.7(3)

170.4(2)

175.5(1)

46.9(3)

7.7(3)

36.0(3)

5.7(3)

170.2(2)

175.5(1)

±

67.2(3)

0.9(3)

45.9(3)

2.0(4)

173.9(2)

172.8(2)

42.3(3)

6.9(4)

30.5(4)

9.7(4)

168.5(2)

172.8(2)

8.5(3)

17.8(3)

±

35.7

2.1

33.1

2.3

132.8

171.2

28.8

0.4

28.5

3.2

134.8

174.3

29.6

24.5

±

C(4) C(3) CO

±

C(3) C(4) CO

21.5(2)

153.1(2)

Æ 45.2

Æ 7.0

C(6) C(5) CO

±

V_av(CC CC)

Æ 43.6

Æ 44.5

Æ 31.1

Æ 46.5

Æ 31.5

V_av(C CC C)

Æ 3.5

Æ 5.2

Æ 3.2

Æ 4.9

Æ 2.0

^a) Torsion angles Vare given for the (pR)-configuration of the homoheptalenes and the (M)-configuration of the heptalene 14 (see

later in the text). ^b) AM1-Calculated parent structure. ^c) Measurement at ambient temperature; data taken from [3]; e.s.d. for V:

ca. 0.48. ^d) Measurement at ca. 125 K; data taken from [4]; e.s.d. for V: ca. 0.2 ± 0.58. ^e) This work; measurements at 173(1) K. ^f) The

X-ray crystal-structure data for 14 (dimethyl heptalene-4,5-dicarboxylate) are taken from [13]: e.s.d. values were not available.

Locants are chosen according to the systematic name of the underlying parent structure bicyclo[5.5.0]dodeca-1,3,5,7,9,11-hexaene,

so that 14 is directly comparable with the homoheptalenes. ^g) It corresponds to the central s-bond (C(1) C(7)) in 14.

1026

Helvetica Chimica Acta ± Vol. 84 (2001)

calculated and X-ray crystal data of 1c manifests that the calculations on the 13,13-X₂-

homoheptalenes are quite reliable and should closely approximate the actual

structures. It is obvious that the increase in bulkiness of the substituents at C(13)

leads to a steady augmentation of W_av(CC C), which is leveled out in the case of 1d

(X Br) and 1e (XMe). These two molecules cannot become flatter in their two p-

halves. The five C-atoms of each half involved (see above) lie almost perfectly in a

plane (cf. Table 4), which also results in nearly identical C X ´´´ C(3/9) and C X ´´´

C(4/10) interatomic distances for 1d and 1e. Further steric pressure by substituents at

C(13) will push V(C(3) C(4) C(5) C(6)), which possesses, for (pR)-configured

homoheptalenes as well as for (M)-heptalenes, normally a negative sign (cf. Tables 3

and 4), to positive values as already indicated by the calculated data of (pR)-13,13-

dimethylhomoheptalene (1e) in Table 4. This change in the sign of

V(C(3) C(4) C(5) C(6)) of appropriately 13,13-disubstituted homoheptalenes

should have a marked influence on the chiroptical properties of such C₂-symmetric

homoheptalenes (see below)⁷).

The UV/VIS spectra of 3 and 13 resemble very much those of comparable

heptalenes (see Figs. 6 and 7 as well as Table 5). The homoheptalene absorption band I,

which is very broad, appears for both diesters at 420 nm and is directly comparable to

the absorption band I of heptalenes (for band definitions, see [8]), which can be

recognized, e.g., as a broad band in the spectra of dimethyl 6,8,10-trimethylheptalene-

Table 4. Comparison of Essential AM1-Calculated Structural Data of 13,13-Disubstituted Homoheptalenes^a)

Homoheptalene (X)

1a (H)

1c (F)^b)

1d (Br)

1e (Me)^c)

Interatomic distances d [pm]

134.5

133.1

135.0

134.6

C(13) X ´´´ C(3/9)

C(13) X ´´´ C(4/10)

255

258

293 (280)

297 (282)

334

325

328 (249)

330 (246)

Bond angles W [8]

134.5

133.1

135.0

134.6

C(1) C(2) C(3)

C(2) C(3) C(4)

C(3) C(4) C(5)

C(4) C(5) C(6)

C(5) C(6) C(7)

124.7

129.5

131.9

133.2

130.0

129.2 (130.2)

131.8 (130.7)

133.0 (132.8)

133.2 (134.8)

131.0 (132.4)

132.8

135.0

134.8

131.9

127.5

133.3

135.6

131.5

126.8

Torsion angles V [8]

134.5

133.1

135.0

134.6

C(1) C(2) C(3) C(4)

C(3) C(4) C(5) C(6)

49.5

37.7

47.7 (40.1)

21.5 ( 22.0)

33.0

0.9

26.4

6.5

Deviation of plane [pm]^d)

134.5

11.4

133.1

135.0

1.5

134.6

1.4

C(2) C(3) C(4) C(5) C(6)

7.5 (7.5)

^a) The torsion angles refer to the (pR)-configuration of the homoheptalenes; 1d: 13,13-dibromo and 1e: 13,13-

dimethyl derivative of 1a. ^b) In parentheses are the data of the X-ray crystal structure of 1c at ca. 125 K [4].

^c) C ´´´ C Distances are given; in parentheses, distances between closest H-atom of H₃C_synC(13), and C(3) or

C(4). ^d) Least-squares deviation of the plane of the indicated fragment.

7

)

For a hypothetical (pR)-13,13-di(tert-butyl)homoheptalene, the AM1 calculation indicates that

V(C(1) C(2) C(3) C(4)) 11.08 becomes smaller than V(C(3) C(4) C(5) C(6)) 20.48. These

values may be regarded as the lower and upper limits, respectively, of the two V of the s-cis-butadiene

subunits in homoheptalenes, the range of which reaches from 678 (13) to 118 and 498 (3) to 208.

Helvetica Chimica Acta ± Vol. 84 (2001)

1027

Fig. 6. UV/VIS Spectra (hexane) of 3 and 13

Fig. 7. UV/VIS Spectrum (hexane) of dimethyl 5,6,8-trimethylheptalene-2,3-dicarboxylate (16)

4,5-dicarboxylate (15) [8] as heptalene analogue of 13, or as a broad shoulder in the

spectra of dimethyl 5,6,8-trimethylheptalene-2,3-dicarboxylate (16) [14] as heptalene

analogue of 3. The homoheptalene absorption band II is visible only as a shoulder at ca.

330 nm in the UV/VIS spectrum of 3, but appears as a clearly separated band at 335 nm

1028

Helvetica Chimica Acta ± Vol. 84 (2001)

Table 5. UV/VIS Spectral Data of the Homoheptalene-dicarboxylates 3 and 13, and of the Heptalene-2,3-

dicarboxylate 16

Compound No.

l_max[nm] (e)

3

13

16

270 (4.50)

279 (4.51)

242 (4.26)

267 (4.31)

330 (sh, 3.65)

±

419 (2.67)

420 (2.70)

334 (3.83)

< 400 (< 2.95)

in 16. Band II is not clearly recognizable in 13, but is most certainly buried under the tail

of the strong homoheptalene absorption band III at 279 nm. Its heptalene analog 15

shows band II as a shoulder at 328 nm and band III also as a shoulder at 280 nm,

followed by a maximum at 261 nm. Bands III appear in 3 (270 nm) and 16 (267 nm) also

at similar wavelengths.

15

E_Me

In summary, one can state that the UV/VIS absorption behaviors of substituted

homoheptalenes and comparable heptalenes are very similar due to the 12e p-

perimeters of both structure types. Nevertheless, the inherent chiral topology of both

c₂-symmetric p-systems is different, and the methano-bridge of 1a and its derivatives

does not allow ring-inversion processes that lead, in heptalenes, from one enantiomer

to the other after double ring-inversion (DRI), principally without involvement of the

DBS process (cf. Scheme 4). A DBS process in substituted heptalenes occurs at lower

temperatures than the DRI process and with retention of a given configuration [15]

(see also [16]). In contrast, homoheptalenes change configuration by the DBS process,

and the interconversion of one enantiomer into the other is realizable only in

symmetrically substituted or at C(4)/C(10)- as well as C(13)-substituted homohepta-

lenes, since the transition state possesses C_s-symmetry for only these structures (cf.

Scheme 4).

The X-ray crystal-structure data in Table 3 demonstrate that the topological

differences between homoheptalenes and heptalenes are caused by a change in the sign

of V at the central s-trans-buta-1,3-diene subunits (C(11/5) C(12/6) C(1/7) C(2/8),

which is negative for (M)-configured heptalenes and positive for (pR)-configured

homoheptalenes. A comparison of the X-ray values of the torsion angles of 3 and the

similarly substituted heptalene-dicarboxylate 16 [14] shows this again quite clearly

(Scheme 5). Moreover, Scheme 5 reveals that the (M)-topology of heptalenes

corresponds with the (pR)-topology of homoheptalenes. As a result of this corre-

spondence in the topology of chirality, the Cotton effects (CE) and their sign in the

long-wavelength region of homoheptalenes and heptalenes with corresponding chiral

topologies should be the same. This is indeed the case as confirmed by the CD spectra

Helvetica Chimica Acta ± Vol. 84 (2001)

1029

Scheme 4

Scheme 5

(pR)-configuration

–36.0(3)°

(M)- = (pR)-configuration

–30.3(3)°

29.9(3)°

46.9(3)°

–124.7(2)°

170.4(2)°

–121.2(2)°

34.5(3)°

170.2(2)°

48.8(2)°

pR

E

pR

E

–36.7(3)°

–48.9(2)°

(M)-16

(pR)-3

of (pS)-3 and (pR)-3 (Fig. 8). The enantiomers of 3 were completely separated on a

semipreparative Chiralcel OD column (see Exper. Part).

Bands I and II of all (M)-configured heptalenes exhibit positive CEs, followed by a

negative CE for heptalene band III (cf. [8][15]). Furthermore, in our hands, all (M)-

configured heptalene-4,5-dicarboxylates or benzo[a]heptalene-6,7-dicarboxylates had

distinctly shorter t_Rvalues than their (P)-antipodes on Chiralcel OD columns with

hexane/EtOH or hexane/i-PrOH (cf. [16]). Based on these observations, there is little

doubt that the enantiomer of 3 with the much shorter t_Rvalue (21 min) and a strong

CE at 425 nm (band I), followed by a second CE, recognizable as a shoulder at

330 nm (band II), must possess the (pR)-configuration (Scheme 5). In turn, the (pS)-

configuration must be ascribed to the enantiomer of 3 with the much larger t_Rvalue

(56 min) and the mirror-image CD spectrum (Fig. 8). However, the CD spectra of the

enantiomers of 3 offer completely independent proof of the absolute configuration of

1030

Helvetica Chimica Acta ± Vol. 84 (2001)

Fig. 8. CD Spectra (hexane) of (pS)- and (pR)-3

the antipodes. According to the exciton-coupling theory (cf. [17]), one expects for the

absorption bands of the chirally twisted MeOCO groups a Davidov splitting, which

should result in opposite signs of their CD in a way that a negative torsion angle

between the two groups would cause a negative CE at longer wavelengths, followed a

positive CE at shorter wavelengths, and vice versa for a positive torsion angle. In other

words, for (pR)-3 with a V(C(3) C(4) C(5) C(6)) value of 48.98 (cf. Scheme 5;

V(OC C(4) C(5) CO) 47.38), one expects first a negative than a positive

CE. Just this situation is met in the CD spectrum of the faster-running antipode of 3, to

which the (pR)-configuration has been ascribed according to the above arguments and

which exhibits, in the spectral region, where the absorption of a,b-unsaturated ester

groups has to be expected, a negative CE at 245 nm and a positive CE at 228 nm.

2.3. Mechanism of the Homoheptalene Formation. The thermal reaction of azulenes

with ADM, which finally results in the formation of heptalene-4,5-dicarboxylates,

leads, in the first step, in a Diels-Alder-type addition reaction, to 1,3a-etheno-1,3a-

dihydroazulenes of type 18 (Scheme 6). These primary intermediates rearrange then

via zwitterionic intermediates and 1,8a-etheno-1,8a-dihydroazulenes of type 19 into the

heptalene-4,5-dicarboxylates (cf. [18]). Such a sequence seems to be impossible or at

least very unlikely for the thermal reaction of homoazulene (2) with ADM due to the

energetically unfavorable first step, which, on steric grounds, could not be realized at

the exo-side of 2, and it would lead to a highly strained tetracyclic intermediate on the

endo-side of 2 (Scheme 6). This conclusion can be drawn from comparison of DH^o_fof

the AM1-calculated basic structures, derived from the hypothetical reaction of ethyne

with azulene (17) and with 2. Moreover, the AM1-calculated DH^o_fvalues of all

conceivable adducts of ethyne and 2 on the exo- and endo-side (Table 6) reveal that all

others are much lower than those of 21 and 22.

Helvetica Chimica Acta ± Vol. 84 (2001)

Scheme 6^a)

1031

1

^a) In parentheses, AM1-calculated DH_f8 values in kcal ´ mol

.

Table 6. AM1-Calculated DH^o_fValues of Possible Primary C₂-Addition Products of Homoazulene (2)^a)

C₂-Species

Ethene

TCNE

Ethyne

ADM

Mode of addition

exo-10,1

endo-10,1

exo-10,2

endo-10,2

exo-10,9

endo-10,9

76.59

104.74

102.85

77.22

80.80

79.74

228.05

255.07

±^b)

232.87

232.65

232.07

123.55

165.93

144.47

110.84

127.44

126.71

38.42

9.78

±^b)

48.84

34.49

35.92

^a) DH^o_fin kcal ´ mol

.

^b) DH^o_fnot calculated.

1

Scott and Kirms observed that tetracyanoethylene (TCNE; ethane-1,1,2,2-tetra-

carbonitrile) reacts with 2 already at 458 in THF with formation of solely the exo-

10,1-adduct (cf. Table 6) [5]. Since TCNE reacts with electron-rich p-systems with

formation of zwitterions (cf. e.g. [19]), they concluded that, under exo-attack, TCNE

and 2 also form a zwitterion composed of homotropylium and dicyanomethanide parts,

which then collapse to the exo-10,1-adduct. Indeed, the exo-10,1-adduct is, according to

1

AM1 calculations (Table 6), by 4.60 kcal ´ mol more stable than the corresponding

exo-10,9-adduct that could also be formed by collapse of the zwitterions. The formation

of an exo-adduct from 2 and TCNE is also in agreement with investigations of

Takahashi et al., who demonstrated that bicyclo[4.3.1]decatetraenide, an anion that is

isoconjugate with 2, exhibits exclusively exo-selectivity on attack of electrophiles such

1032

Helvetica Chimica Acta ± Vol. 84 (2001)

as deuteron, CO₂or alkyl and silyl halides [20]. CNDO/2 Calculations of these authors

were in agreement with a preferred exo-attack of electrophiles due to outward-bent

orbital lobes at C(7) and C(9) on the exo-face and, in turn, inward-bent orbital lobes on

the endo-face in the HOMO of the anion, which possesses a nodal plane passing

through C(8) and the opposite C(3),C(4) bond, thereby causing less bonding

interaction with the LUMO of the electrophiles on its endo-face. Therefore, it can be

assumed that the same exo-selectivity will be exhibited by 2 as has already been

proposed by Scott and Kirms on the basis of their result with 2 and TCNE. This means

that the electron-poor ADM should attack 2 also exclusively on the exo-side under

formation of corresponding zwitterions exo-23 (Scheme 7), comprising a homotropyl-

ium and an allenolate part. The break-down of these zwitterions will give the two

cyclobutene exo-10,1- and exo-10,9 intermediates exo-24 and exo-25, respectively,

which, upon ring opening, form the observed homoheptalenes 13 and 3 (Scheme 7).

However, the ratio for 3/13 of 4 :1 determined is not in agreement with the AM1-

calculated DH^o_fvalues for exo-24 and exo-25 (cf. Table 6), which strongly accounts for

the almost exclusive formation of exo-24 and thus 13 (DDH^o_f 3.93 kcal ´ mol

1

corresponds to ca. 99.6% of 13 and 0.4% of 3 at 828, assuming DDS^o_fꢁ 0), in analogy

to the result of the reaction of 2 and TCNE. This discrepancy confirms that exo-23 does

not represent a relaxed zwitterion with internal rotational freedom. We, therefore,

assume that the addition of ADM to 2 takes place via a charge-compensated transition

state, leading to exo-23 in a conformation that kinetically favors the formation of the

exo-10,9-intermediate 25, from which the homoheptalene-dicarboxylate 3 is derived.

Indeed, when we carried out the reaction of 2 and ADM in boiling MeCN as a stronger

solvating, polar solvent, we observed the formation of 3 and 13 in a ratio of 3 :2,

consistent with a larger internal rotational freedom of exo-23 due to its better

stabilization by solvation.

We believe we can exclude the possibility that 2 is subject to an exo- and endo-

attack of ADM, which, in addition to exo-23, will also give endo-23. Collapse of endo-23

could lead only to the endo-10,9-intermediate 25, since its endo-10,1-counterpart would

be extremely strained (cf. DH^o_fin Table 6). This means that 3 would mainly arise from

the endo-attack, and 13 from the exo-attack of ADM on 2. However, the AM1-

calculated DH^o_fvalues in Table 6 indicate that, in such a case, there would be an

energetically much better path for a collapse of endo-23, namely to form the tetracyclic

endo-10,2-product 26, which is favored over endo-25 by almost 13 kcal ´ mol ¹according

to their AM1-calculated DH^o_fvalue (cf. Table 6)⁸).

2.4. p-Substituted Derivatives of 3. On the basis of earlier results on electrophilic

substitution reactions of 2 by Scott et al. [23], we subjected 2 to the Vilsmeyer

formylation procedure with N,N-dimethylformamide (DMF) and POCl₃at 08. After

workup and chromatography on silica gel, the pure carbaldehyde 27 was obtained as an

unstable red oil in a yield of 67% (Scheme 8). A control reaction with azulene (17)

under identical conditions gave azulene-1-carbaldehyde in a yield of 96%, which

demonstrates that some homoazulene-8-carbaldehyde (27) may have been destroyed

8

)

The cyclization of endo-23 to endo-26 is completely analogous to the exclusive 1,8-addition of ADM to

azulenes under Lewis acid catalysis [14] [21][22], in contrast to the purely thermal reaction that follows the

path shown in Scheme 7.

Scheme 7

OMe

O

E

δ

E

C

+

3

ADM

E

H

2

exo-23

exo-25

3/13 =

4:1 (DCE)

?

3:2 (MeCN)

H

13

O

H

E

C

H

O

C

H

E

MeO

E

MeO

exo-24

endo-23

endo-26

exo-23

endo-24

1034

Helvetica Chimica Acta ± Vol. 84 (2001)

Scheme 8

a) 1. POCl₃(1.2 equiv.)/DMF, 08; 2. 1 equiv. 2 in DMF, 08/15 min; 67%. b) 1. [Ph₃PCH₂Ph] Br (3.24 equiv.)/t-

BuOK (3.43 equiv.)/THF, 208 ! 788. 2. 1 equiv. 15 in THF, 788/15 min ! 208/1 h; 99%. c) 1 equiv. (E)/(Z)-

28/ClCH₂CH₂Cl and 2.64 equiv. ADM, 828/18 h; 42%. d) 1 equiv. (Z)-29/toluene and 1.8 equiv. of 1.5m

DIBAH/toluene, 908/4 h; 28% of (Z)-31 and 19% of (Z)-32. e) On standing in an ice-box for 1 h; quantitative.

already during workup and chromatography (cf. [23]). Whereas the reduction of the

aldehyde function of 27 to a Me group with NaBH₄/BF₃´ Et₂O in diglyme/Et₂O (cf.

[24]), a procedure that is very successful in the case of azulene-1-carbaldehydes (cf.,

e.g., [18a][25]), failed since 2 was destroyed under the reaction conditions, we were

able to perform a Wittig reaction with 2 and benzylidene(triphenyl)l⁵-phosphane in

THFat 78 to 08. The corresponding 8-styrylhomoheptalene 28 was obtained in almost

quantitative yield as a 1.6 :1 (Z)/(E)-mixture. The two isomers could be isolated by

chromatography on alumina with hexane as eluant as dark red oils, which have been

fully characterized spectroscopically (cf. Exper. Part). When the 1.6 :1 mixture (Z)/

(E)-28 was reacted with ADM in boiling ClCH₂CH₂Cl, the 2-styrylhomoheptalene-4,5-

dicarboxylate 29 was obtained in 42% yield as a 5 :1 (Z)/(E)-mixture. The (Z)- and

(E)-isomers of the corresponding 2,3-dicarboxylate 30 could not be detected in the

reaction mixture, but they might have been formed and destroyed under the reaction

conditions. The low yield of 29 in comparison to that of the reaction of 2 and ADM,

Helvetica Chimica Acta ± Vol. 84 (2001)

1035

and the change of the (Z)/(E) ratio in the course of the reaction reveal that the

formation of (Z)-29 and (E)-29 must have been accompanied by partial destruction of

the homoazulenes (Z)-28 and (E)-28. Nevertheless, by careful and repeated

chromatography on silica gel of the 5 :1 mixture of (Z)/(E)-29 could be separated,

whereby (Z)-29 was obtained as a brownish red oil and (E)-29 as brown-red needles.

The two isomers were fully characterized spectroscopically (cf. Exper. Part). Their UV/

VIS spectra in hexane are displayed in Fig. 9. In comparison with the spectrum of 3

(Fig. 6), a bathochromic shift of 30 nm of band I, accompanied by a slight hyper-

chromism, is observed for both isomers. Band II is completely buried under the

bathochromically shifted intense band III of (E)-29, due to conjugative interaction of

the homoheptalene chromophore with the (E)-styryl moiety at C(2). However, (Z)-29

displays band II in comparison with 3 as a strongly enhanced shoulder at 310 nm, since

band III of (Z)-29, which appears at 267 nm (270 nm for 3), is not markedly influenced

by conjugation with the (Z)-configured styryl residue at C(2). It forms, on steric

grounds, with the homoheptalene p-system a torsion angle of 74.28, whereby the

torsion angle within the styryl group is 33.78 (AM1 calculation of the energetically most

relaxed conformer). The corresponding AM1 calculations provided, for (E)-29, torsion

angles of 162.88 and 22.88, respectively.

The reduction of (Z)-29 with DIBAH in toluene gave the corresponding

homoheptalene-4,5-dimethanol (Z)-31 together with some of the homoheptaleno-

furan-1-one (Z)-32 (Scheme 8). On standing, (Z)-31 isomerized completely to (E)-31.

Its UV/VIS spectrum in hexane is displayed in Fig. 10. The homoheptalene bands I and

II are well-recognizable as shoulders at ca. 430 and 362 nm with an enhanced intensity

of band I in comparison to that of the corresponding dicarboxylate (E)-29, which

Fig. 9. UV/VIS Spectra (hexane) of dimethyl 1-[(E)- and (Z)-2-phenylethen-1-yl]homoheptalene-4,5-dicarbox-

ylate ((E) and (Z)-29)

1036

Helvetica Chimica Acta ± Vol. 84 (2001)

Fig. 10. UV/VIS Spectrum (hexane) of 1-[(E)-2-phenylethen-1-yl]homoheptalene-4,5-dimethanol ((Z)-31)

appears as a broad maximum at 450 nm (see Fig. 6). The absorption band of the (E)-

styryl-substituted homoheptalene chromophor of (E)-31 and (E)-29 emerges at almost

the same wavelength in both compounds, securing the assignment of these bands (see

also the UV/VIS spectrum of (Z)-29 in Fig. 6, where this band is not present).

Unfortunately, the spectra of none of the new unsymmetrically substituted

homoheptanes showed any indication of their DBS isomers. The AM1-calculated

DH^o_fvalues of the (E)- and (Z)-isomers of 29, and their DBS forms 29' as well as those

of the corresponding, but not observed (E)- and (Z)-isomers of the homoheptalene-

2,3-dicarboxylate 30 and their DBS forms 30' are listed in Table 7. In Table 7 are also

included the calculated DH^o_fvalues of 3 and 13 and the corresponding Me-substituted

dicarboxylates and their DBS forms. All values indicate that the energetically favored

Table 7. AM1-Calculated DH^o_fValues of the DBS Isomers of Vicinal Dimethyl Homoheptalene-dicarboxylates^a)

R

H

Me

(E)-Styryl

(Z)-Styryl

Position of R

and E_Me

C(6), C(2,3)

C(2), C(5,6)

DDH^o_f

64.10

59.86

4.24

69.18

66.14

3.04

22.49

19.32

3.17

19.69

17.03

2.68

C(2), C(4,5)

C(6), C(3,4)

DDH^o_f

63.92

62.05

1.87

70.34

67.07

3.27

23.58

20.64

2.94

20.30

17.64

2.66

o

^a) DH_fin kcal ´ mol for the energetically most relaxed conformers.

1

Helvetica Chimica Acta ± Vol. 84 (2001)

1037

DBS forms are those that are comparable with the isolated and characterized isomers 3

and 13 and (Z)- and (E)-29. Therefore, according to AM1 calculations, we had no

opportunity to observe their DBS forms in thermal equilibrium. The values also

indicate that the C(2)-substituted homoheptalene-dicarboxylates, derived from 3,

exhibit increased DDH^o_fvalues, whereas the C(6)-substituted homoheptalene-dicar-

boxylates that follow from 13 show slightly decreased DDH^o_fvalues. Both homohepta-

lene-dicarboxylate series approach, having similar DDH^o_fvalues of ca. 3 kcal ´ mol

1

consistent with the 2-methylhomoheptalene being more stable by ca. 1 kcal ´ mol ¹than

its DBS isomer, 6-methylhomoheptalene, on the basis of AM1 calculations.

We will continue our work with the aim to gain better understanding of the nature

of the exited-state of the inherently chiral p-system of the c₂-twisted heptalene and

homoheptalene s-skeletons.

We thank Dr. A. Linden for the X-ray crystal-structure determinations, our NMR laboratory, in particular

Nadja Walch, Martin Binder, and Dr. Gudrun Hopp, for specific NMR measurements, Prof. M. Hesse and his

group for mass-spectrometric measurements, and, finally, Peter Uebelhart for the chromatographic resolution of

the homoheptalene-dicarboxylate. The financial support of this work by the Swiss National Science Foundation

is gratefully acknowledged.

Experimental Part

General. TLC: glass plates covered with silica gel 60 F₂₅₄(Merck) or on aluminium foils covered with Alox

60 F₂₅₄, type E, 0.2 mm (Merck). Column chromatography (CC): silica gel 60 (Merck), grain size 0.040 ±

0.063 mm or on alumina (Fluka), type 5016 A basic; act. IV. M.p. (not corrected): Mettler FP 5/52. UV/VIS

Spectra: were recorded on a Perkin-Elmer Lambda19 spectrometer; l (nm), in parentheses log e. IR Spectra:

Perkin-Elmer 1600 Series FT-IR spectrometer; wave numbers in cm ¹; characterization of the band intensities

(transmission): 0 ± 20% vs. 20 ± 40% s, 40 ± 60% m, 60 ± 80% w, 80 ± 100% vw. NMR Spectra: Bruker ARX-300

(300/75 MHz), Avance DRX-500 (500/125 MHz), and Avance DRX-600 (600/150 MHz); chemical shifts (d,

ppm) relative to CD(H)Cl₃(7.27/77.00 ppm); for complete assignments of ¹H-NMR signals COSY, TOCSY,

NOESY, ROESY 2D- or 1D-NMR methods were applied; for complete assignments of ¹³C-NMR signals

HMBC and HSQC 2D-NMR methods were employed. If not stated otherwise, the spectra were recorded at 300

and 75 MHz, resp., in CDCl₃. MS: Varian SSQ 700; Ionization by EI (70 eV) or Cl (NH₃): m/z, rel. intensities

(%). GC/MS: Hewlett Packard HP-5971 Series (mass-selective detector; EI, 70 eV) and HP-5890 Series II

(GC); carrier gas He; WCOT capillary column HP-5; 25 m Â 0.3 mm, 0.2 mm; injector temp. 2808, starting temp.

1008 for 7 min, temp. rate 208/min, final temp. 2408.

1. Homoazulene and Derivatives. ± 1.1. Bicyclo[5.3.1]undeca-1,3,5,7,9-pentaene (Homoazulene; 2, cf. [2]).

1.1.1. 1-Diazo-4-phenylbutan-2-one (4). In a 2-neck flask I, connected via a condenser with a 3-neck flask II,

finely powdered KOH (81.0 g, 1.44 mol) was added under ice-cooling to a mixture of carbitol (240 ml), H₂O

(120 ml), and Et₂O (80 ml). Flask II contained a soln. of dihydrocinnamoyl chloride (Fluka; 71.62 g, 0.425 mol)

in Et₂O (100 ml), which was kept by cooling at 108. Flask I was heated to 558 and, after most of the Et₂O had

been distilled into flask II, a soln. of ꢀdiazaldꢁ (N-methyl-N-nitroso-p-toluene sulfonamide, Fluka; 270.0 g,

1.26 mol) in a mixture of Et₂O (500 ml) and THF (250 ml) was dropped into the mixture of flask I within 2 h,

whereby the temp. was maintained at 558, so that the formed CH₂N₂was continuously co-distilled with Et₂O into

flask II. After all of the soln. of ꢀdiazaldꢁ had been added to flask I, Et₂O (50 ml) was added to flask I and

distilled into flask II until the distilling Et₂O became colorless. Flask II was brought within 2 h at r.t. and then

heated for an additional h at 508. The excess CH₂N₂in flask II that distilled off with Et₂O was destroyed by

reaction with AcOH. The residual oil was diluted with toluene (ca. 50 ml), which was again distilled off. The

residue was then completely freed under h.v. from traces of solvents to give 4 (74.04 g, 100%) as a yellow oil. A

sample for analysis was purified by CC (silica gel; hexane/Et₂O 1:1). R_f(hexane/Et₂O 1:2) 0.28. UV (hexane):

l_max246 (4.12); l_min222 (3.60). IR (film): 2105vs (CHN₂), 1640vs (CO). ¹H-NMR: 7.28, 7.19 (5 arom. H); 5.17

(s, H C(1)); 2.95 (t, J 7.7, CH₂(4)); 2.63 (br. t, J 7.4, CH₂(3)). ¹³C-NMR: 193.76 (s, C(2)); 140.53 (s, C(1) of

1038

Helvetica Chimica Acta ± Vol. 84 (2001)

Ph); 128.43, 128.21 (2d, C_oand C_mof Ph)); 126.15 (d, C_pof Ph); 54.45 (t, C(4)); 42.26 (t, C(3)); 30.82 (d, C(1)).

.

GC/MS (C₁₀H₁₀N₂O; 174.20): 201 (36), 173 (6, [M 1] ), 146 (13, [M N₂] ), 118 (15), 115 (8), 91 (56), 65

(14), 55 (100).

1.1.2. 3,4-Dihydro-2H-azulen-1-one (5). In a 750-ml 3-neck flask, equipped with an efficient condenser,

thermometer, and N₂inlet, Rh₂(AcO)₄(0.052 g, 0.12 mmol) was dissolved in CH₂Cl₂(430 ml). The blue-green

soln. was heated at reflux (428), and, under N₂, a soln. of 4 in CH₂Cl₂(10 ml) was added within 15 h via a syringe,

driven by a dosage pump, whereby the tip of the elongated hollow needle was placed in the refluxing solvent

stream at the condenser. As a result of the very slow addition of 4 and the additional dilution by the refluxing

solvent, an optimally high dilution effect for the catalyzed intramolecular Buchner reaction could be attained.

After the addition of 4, the soln. was heated for a further h and then cooled to r.t. A small amount of Et₃N

(0.2 ml) was added, which caused a transitory darkening from yellow to brown and a spontaneous warming of

the soln. to 408. Finally, the initial yellow color returned. After 0.5 h, the soln. was filtered over silica gel, CH₂Cl₂

was distilled off, and the yellow residue was subjected to CC (silica gel; hexane/t-BuOMe 5 :1). The slightly

greenish yellow oil obtained was crystallized from hexane/t-BuOMe/20 :1 to give pure 5 (21.87 g, 69%).

Colorless needles. M.p. 298. R_f(hexane/t-BuOMe 1:1) 0.29. UV/VIS (hexane): l_max226 (4.14),267 (3.54); l_min

247 (3.37). IR (film): 3378vw, 3023vs, 2958s, 2922s, 2864m, 2829s, 1695vs, 1622vs, 1561m, 1430vs, 1406s, 1384s,

1334vs, 1276vs, 1216m, 1140vs, 1078m, 1005s, 988s, 897w, 822s, 794s, 737vs, 697s, 660s, 622vs, 520m. ¹H-NMR: 6.75

(d, ³J(8,7) 11.1, H C(8)); 6.53 (dd, ³J(7,8) 11.1, ³J(7,6) 5.8, H C(7)); 6.15 (dd, ³J(6,5) 9.9, ³J(6,7) 5.8,

H

C(6)); 3.34 (dt, ³J(5,6) 9.9, ³J(5,4) 6.3, H C(5)); 2.85 (d, ³J(4,5) 6.3, CH₂(4)); 2.72 (t-like, ³J(2,3)

4.4, CH₂(2)); 2.52 (m_c, CH₂(3)). ¹³C-NMR: 205.43 (s, C(1)); 165.88 (s, C(3a)); 136.98 (s, C(7a)); 130.67, 128.62,

122.44, 120.99 (4d, C(5,6,7,8)); 35.62 (t, C(2)); 31.12 (t, C(4)); 29.77 (t, C(3)). GC/MS (C₁₀H₁₀O; 146.19): 146

.

(64, M ), 131 (5), 117 (69), 115 (53), 104 (100), 91 (20), 78 (16), 63 (31).

1.1.3. Tricyclo[5.3.1.0^1,7]undeca-2,4-dien-10-one (6). NaH (2.172 g, 90.24 mmol), obtained from the

dispersion of NaH in oil, was added to dry DMSO (320 ml). Under intense stirring, trimethylsulfoxonium

iodide (Fluka; 36.13 g, 164.2 mmol) was added within 5 min. The evolution of H₂was indicated by transitory

foaming. After additional stirring for 30 min, a soln. of 5 (12.00 g, 82.08 mmol) in dry DMSO (70 ml) was added

dropwise within 40 min. The resulting dark brown mixture was heated for 1 h at 758, whereby the color changed

to light yellow. The cooled mixture was poured onto ice and extracted with CH₂Cl₂. After filtration over cotton

wool, the solvent was distilled off, and the residue was subjected to CC (silica gel; hexane/Et₂O 2 :1) to give pure

6 (9.60 g, 73%). Slightly yellow oil. R_f(hexane/Et₂O 2 :1) 0.26. UV/VIS (hexane): l_max230 (3.56), 266 (3.60);

l_min211 (3.41), 248 (3.52). IR (film): 3023m, 2935m, 2869m, 2835w, 1721vs, 1607m, 1435m, 1392w, 1348w, 1297s,

1262m, 1176w, 1088m, 1055s, 1019m, 924w, 851m, 820w, 779m, 707vs, 654w, 603m, 547w. ¹H-NMR: 6.47

(d, ³J(2,3) 11.8, H C(2)); 5.86 (ddd, ³J(4,5) 11.1, ³J(4,3) 6.0, ⁴J 2.0, H C(4)); 5.63 (dd, ³J(3,2) 11.8,

³J(3,4) 6.0, H C(3)); 5.61 ± 5.53 (m,H C(5)); 2.69 (complex ABX system, d_Ad_Bꢁ 8, CH₂(6)); 2.58

(d, ²J(11_b,11_a) 4.4, H_bC(11)); 2.19 ± 2.01 (m, CH₂(8,9)); 1.25 (dd, ²J(11_a,11_b) 4.4, ⁴J 0.9, H_aC(11)).

¹³C-NMR: 211.92 (s, C(10)); 128.68, 127.43, 127.18, 122.78 (4d, C(2,3,4,5)); 58.06 (s, C(1)); 41.30 (s, C(7)); 32.47,

.

31.93, 28.14 (3t, C(6,8,9)); 22.60 (t, C(11)). GC/MS (C₁₁H₁₂O; 160.22): 160 (45, M ), 145 (8), 133 (35), 117

(100), 104 (80), 91 (63), 78 (61), 77 (60).

1.1.4. Tricyclo[5.3.1.0^1,7]undeca-2,4,9-triene (8). 1.1.4.1. Tricyclo[5.3.1.0^1,7]undeca-2,4-dien-10-one Tosylhy-

drazone (7). Ketone 6 (8.63 g, 53.86 mmol) was dissolved in THF (13 ml) and TsNHNH₂(10.64 g, 57.13 mmol)

was added in portions at r.t. Stirring was continued for 15 h at r.t., and the pale yellow suspension obtained was

filtered and washed with a small amount of ice-cold Et₂O. The now colorless powder of 7 was dried in h.v.

(15.37 g, 87%). M.p. 190 ± 1928 ([2]: 190 ± 1928). R_f(hexane/Et₂O 2 :1) 0.13. UV (MeCN): l_max226 (3.95); l_min

212 (3.87). IR (KBr): 3216vs, 3025m, 2933m, 1654m, 1599w, 1494w, 1436m, 1400s, 1343vs, 1310m, 1168vs, 1076s,

1023s, 927m, 850w, 816s, 710vs, 697s, 662s, 622m, 567vs, 549s. ¹H-NMR: 7.88 (d, ³J 8.4, H_oof Ph); 7.32 (d-like,

3

4

³J 8.4, H_mof Ph); 6.48 (d, ³J(2,3) 12.0, H C(2)); 5.82 (ddd, ³J(4,5) 11.3, J(4,3) 6.0, J 2.2, H C(4));

5.56 (dd, ³J(3,2) 12.0, ³J(3,4) 6.0, H C(3)); 5.52 ± 5.47 (m, H C(5)); 2.59 (complex ABX system, d_Ad_Bꢁ

8 Hz); 2.43 (s, Me Ph)); 2.36 (d, ²J(11_a,11_b) 4.6, H_aC(11)); 2.07 ± 1.75 (m, CH₂(8,9)); 0.80 (dd, ²J(11_b,11_a)

4.6, ⁴J 0.7, H_bC(11)). ¹³C-NMR: 167.75 (s, C(10)); 143.82 (s, C_pof Ph)); 135.29 (s, C_ipsoof Ph)); 129.75, 128.67,

127.61, 121.65 (4d, C(2,3,4,5)); 129.38 (d, C_mof Ph)); 128.08 (d, C_oof Ph)); 56.01 (s, C(1)); 37.01 (s, C(7)); 31.99,

.

30.87, 23.09 (3t, C(6,8,9)); 21.49 (q, Me Ph); 20.96 (t, C(11)). EI-MS (C₁₈H₂₀N₂O₂S; 328.44): 328 (14, M ), 173

.

(100, [M Ts] ), 144 (27, [M TsN₂H] ), 129 (96), 115 (32), 91 (68), 77 (18).

1.1.4.2. Formation of 8. A 1.45m soln. of MeLi in Et₂O (188 ml, 0.273 mol) was added at r.t. within 1 h to a

suspension of 7 (43.38 g, 0.132 mol) in Et₂O (860 ml). The evolution of CH₄was indicated by foaming of the

stirred mixture. The initially colorless mixture became yellow after addition of the first mol-equiv. of MeLi, and

turned orange after the second mol-equiv. of MeLi. The elimination of N₂was completed by stirring overnight at

Helvetica Chimica Acta ± Vol. 84 (2001)

1039

r.t., whereby the mixture turned brown. H₂O (10 ml) was added under ice cooling, whereby the temp. of the

mixture rose for a short time to 558. Et₂O (200 ml) was added, and the mixture was 3 times washed with ice-cold

H₂O. The waste H₂O was re-extracted with Et₂O (100 ml), and the combined Et₂O extracts were dried

(MgSO₄). The light orange-colored residual oil of the Et₂O extracts was subjected to CC (silica gel; pentane) to

give 8 as a colorless volatile oil (17.30 g, 91%). R_f(pentane) 0.68. UV (hexane): l_max221 (sh, 3.61), 279 (3.44);

l_min251 (3.30). IR (film): 3017vs, 2905vs, 2832vs, 1600s, 1436m, 1287w, 1253m, 1195m, 1143w, 1122w, 1075m,

1034m, 938s, 902m, 848m, 805m, 740vs, 706s, 685vs, 623vs, 556m, 512s. ¹H-NMR (600 MHz, CDCl₃): 6.19

(d, ³J(2,3) 11.6, H C(2)); 5.89 (dtd, ³J(10,9) 5.5, ⁴J(10,8) 1.8, ⁴J(10, H_aC(11)) 0.5, H C(10)); 5.80

(ddd, ³J(4,5) 11.2, ³J(4,3) 6.0, ⁴J 2.6, H C(4)); 5.55 (ddd, ³J(5,4) 11.2, ³J(5,6_exo) 8.4, ³J(5,6_endo) 4.4,

H

C(5)); 5.51 (dd, ³J(3,2) 11.6, ³J(3,4) 6.0, H C(3)); 5.30 (dt, ³J(9,10) 5.5, ³J(9,8) 2.2, H C(9)); 2.68

4

(A of ABX, ²J_AB 14.8, ³J(6,5) 8.3, H_exoC(6)); 2.64 (A of ABXY, ²J_AB 17.4, ³J(8,9) ꢁ J(8, H_aC(11)) 2.2,

2

H_endoC(8)); 2.60 (B part of the ABX system, br., J_AB 14.8, H_endoC(6)); 2.39 (B of ABXY, J_AB 17.4,

³J(8,9) ꢁ J(8, H_aC(11)) 2.2, H_exoC(8)); 2.35 (d, ²J(11_a,11_b) 3.4, H_aC(11)); 0.34 (d, ²J(11_b,11_a) 3.3,

4

H_bC(11)). ¹³C-NMR (150 MHz, CDCl₃): 138.47 (d, C(10)); 134.75 (d, C(2)); 128.68 (d, C(4)); 128.49

(d, C(5)); 125.65 (d, C(9)); 121.80 (d, C(3)); 52.94 (s, C(7)); 43.50 (t, C(8)); 38.65 (s, C(1)); 32.48 (t, C(6)); 24.90

.

(t, C(11)). GC/MS (C₁₁H₁₂, 144.22): 144 (92, M ), 141 (78), 128 (99, C₁₀H₈), 115 (100), 103 (71), 91 (80), 77

(79).

1.1.5. Bicyclo[5.3.1]undeca-2,4,9-triene-1,7-diyl Diacetate (9). General. As mentioned already in the Theor.

Part, the formation of 9 had been, in our hands, the most delicate step in the synthesis of 2. Neither by varying

the reaction conditions (cf. Table 1) nor by changing the batch of Pb(OAc)₄(Fluka) could we attain the

yield that had been reported by Scott and Kirms [5]. We describe here our run that delivered the highest yield.

Formation of 9. To a mixture of benzene (270 ml) and glacial AcOH (110 ml) were added, at 158, 8 (5.98 g,

41.46 mmol) and Pb(AcO)₄. The mixture was stirred during 1 h at 158, whereby the initially colorless suspension

turned yellow. Stirring was continued at r.t. during 15 h, and then a further amount of Pb(AcO)₄(18 g, 40 mmol)

was added. The run was terminated after 3 h stirring at 358. Most of the solvent mixture was distilled off, and the

remaining yellow suspension was filtered. The residue of the filtrate was passed through a short column (silica

gel; hexane/t-BuOMe 4 :1) and washed several times with H₂O to remove the last amounts of Pb(AcO)₂and

then with NaHCO₃and sat. NaCl soln. After drying (MgSO₄), the residue of the org. phase was dissolved in

toluene (50 ml), which was removed again by distillation. Finally, CC of the residue (silica gel; hexane/t-BuOMe

8 :1) led to a colorless oil, which crystallized after a short time of standing at r.t. Recrystallization (hexane/

TBME 10 :1) gave pure 9 (3.05 g, 28%) as colorless crystals. M.p. 92 ± 948. R_f(hexane/TBME 6 :1) 0.34. UV

(hexane): l_max247 (3.63); l_min223 (3.33). IR (KBr): 2937w, 1726vs, 1434m, 1372vs, 1348m, 1330w, 1259vs,

1229vs, 1195s, 1144w, 1070s, 1049s, 1023s, 983m, 959vs, 943s, 915m, 880s, 828m, 793s, 763m, 744s, 724w, 689w,

662w, 632m, 605w, 575w. ¹H-NMR (600 MHz, CDCl₃): 6.15 (dd, ³J(4,5) 10.8, ³J(4,3) 4.8, H C(4)); 5.79 ±

5.77 (m, H C(9)); 5.77 (dd, ³J(3,2) 12.8, ³J(3,4) 4.9, H C(3)); 5.70 ± 5.66 (m,H C(5), H C(10)); 5.46

(dd, ³J(2,3) 12.5, ⁴J 1.8, H C(2)); 3.51 (d, ²J(11_a,11_b) 12.8, H_aC(11)); 2.86 (br. t-like, J 2.6, CH₂(8));

2

3

2

3

2.82 (ddd, A of ABX, J_AB 13.8, J_AX 7.0, ⁴J 1.5, CH₂(6)); 2.77 (dd, B of ABX, J_AB 13.8, J_BX 9.1,

CH₂(6)); 2.30 (dt, ²J(11_b,11_a) 12.8, ⁴J 1.7, H_bC(11)); 1.984, 1.977 (2s, 2 MeOCO). ¹³C-NMR (150 MHz,

CDCl₃): 170.43, 169.67 (2s, 2 MeOCO); 130.60 (d, C(2)); 129.92, 129.90 (2d, C(4,10)); 128.46 (d, C(5)); 127.18

(d, C(9)); 125.24 (d, C(3)); 84.02, 83.86 (2s, C(1,7)); 36.75, 36.74 (2t, C(8,11)); 33.47 (t, C(6)); 22.44, 22.39 (2q,

.

MeOCO C(1,7)). GC/MS: 262 (15, M ), 202 (5, [M AcOH] ), 160 (100, [M H (AcOH MeCO] ),

.

142 (55, [M 2 AcOH] ), 131 (33), 117 (41), 115 (21), 104 (8), 91 (27), 77 (9). Anal. calc. for C₁₅H₁₈O₄

(262.31): C 68.68, H 6.92; found: C 68.77, H 6.95.

In addition, the structure of 9 was established by an X-ray crystal-structure analysis (cf. Fig. 2 and Table 9).

1.1.6. Formation of 2. Since the last intermediates on the way to 2 were quite unstable, the following steps

were performed one after the other.

1.1.6.1. Bicyclo[5.3.1]undeca-2,4,8,10-tetraen-7-yl Acetate (10). To a soln. of 9 (5.03 g, 19.18 mmol) in

heptane (250 ml) were added under stirring at r.t. in the following order: Pd(OAc)₂(0.077 g, 0.34 mmol), Ph₃P

(0.56 g, 2.14 mmol), and Na₂CO₃(1.87 g, 12.4 mmol). The slightly yellow mixture was stirred at 858 during 15 h,

which led to a light yellow suspension. After filtration over Celite, most of the solvent was distilled off. The

residue was subjected to CC (silica gel; hexane/t-BuOMe 50 :1) to give pure 10 (3.71 g, 96%). Unstable, bright

yellow oil. R_f(hexane/t-BuOMe 20 :1) 0.47. IR (film): 3014s, 2935m, 1737vs, 1550w, 1449s, 1369vs, 1282s, 1234vs,

1131w, 1052vs, 1020s, 985m, 859m, 830w, 746s, 709s, 641m, 606m, 579w, 557w, 497w. ¹H-NMR: 6.10 ± 6.06

3

4

3

(m, H C(2,8)); 5.90 (ddd, ³J(4,5) 12.1, J(4,3) 6.4, J 1.8, H C(4)); 5.83 (dd, ³J(9,8) 9.3, J(9,10) 4.4,

2

H

C(9)); 5.59 (d, ³J(10,9) 4.4, H C(10)); 5.51 ± 5.44 (m, H C(3,5)); 4.04 (ddd, A of ABX, J_AB 14.1,

1040

Helvetica Chimica Acta ± Vol. 84 (2001)

2

3

³J_AX 8.5, ⁴J 1.3, H_endoC(6))⁹); 3.57 (s, CH₂(11)); 2.38 (dd, B of ABX, J_AB 14.1, J_BX 7.9, H_endoC(6));

2.04 (s, MeOCO). ¹³C-NMR: 170.51 (s, MeOCO); 138.95 (s, C(1)); 136.10, 131.13, 128.97, 127.78, 126.04, 125.00,

122.92 (7d, C(2) ± C(5), C(8) ± C(10)); 84.71 (s, C(7)); 34.42, 31.73 (2t, C(6,11)); 21.95 (q, MeOCO). GC/MS

.

(C₁₃H₁₄O₂; 202.26): 202 (33, M ), 160 (82), 159 (78, [M MeCO] ), 145 (100), 142 (79, [M AcOH] ), 141

(92), 127 (93), 115 (96), 103 (21), 91 (3), 77 (73).

1.1.6.2. Bicyclo[5.3.1]undeca-2,4,8,10-tetraene-7-ol (11). To a soln. of freshly prepared 10 (3.71 g,

18.34 mmol) in Et₂O (80 ml) was dropped at r.t. and within 30 min a 1.45m soln. of MeLi in Et₂O (26.2ml,

38 mmol). Stirring was continued for 1 h. The now brown suspension was cooled with ice, and H₂O (ca. 50 ml)

was added. The aq. phase was extracted with Et₂O (3Â), and the combined org. layers were washed with sat.

NaCl soln. After drying (Na₂SO₄) and removal of the solvent, the residue was further dried and freed from

traces of solvent in h.v. Alcohol 11 was obtained as a pale yellow, unstable oil (2.64 g, 90%). A sample of 11 was

further purified by chromatography (alumina; hexane/t-BuOMe 2 :1). R_f(hexane/t-BuOMe 2 :1) 0.21. IR

(film): 3359vs, 3013vs, 2973vs, 1713vw, 1589m, 1548w, 1477s, 1374vs, 1278s, 1231s, 1204s, 1083vs, 1015s, 988m,

926s, 871m, 857s, 830w, 740vs, 708vs, 653m, 613m, 559s, 478vw. ¹H-NMR: 6.08 (d, ³J(2,3) 11.8, H C(2)); 5.89

(ddd, ³J(4,5) 12.2, ³J(4,3) 6.3, ⁴J 1.7, H C(4)); 5.83 (dd, ³J(9,8) 9.4, ³J(9,10) 4.9, H C(9)); 5.73 (d-

like, ³J 8.4, H C(8)); 5.59 ± 5.52 (m, H C(5, 10)); 5.46 (dd, ³J(3,2) 11.8, ³J(3,4) 6.3, H C(3)); 4.04

2

3

2

(dd, A of ABX, J_AB 13.6, J_AX 8.4, H_endoC(6)); 3.62, 3.07 (2d, AB, J_AB 11.9, CH₂(11)); 1.88 (ddd, B of

ABX, ²J_AB 13.6, ³J_BX 8.0, ⁴J 1.7, H_exoC(6)); 1.67 (s, OH). ¹³C-NMR: 140.24, 131.13, 128.99, 128.17, 125.65,

125.44, 123.04 (7d, C(2 ± 5), C(8 ± 10)); 139.49 (s, C(1)); 76.56 (s, C(7)); 37.88, 35.08 (2t, C(6,11)). GC/MS

.

(C₁₁H₁₂O; 160.22): 160 (77, M ), 159 (74, [M H] ), 145 (92), 142 (30, [M H₂O] ), 131 (81), 127 (89), 115

(100), 103 (71), 91 (80), 77 (81).

1.1.6.3. Bicyclo[5.3.1]undeca-2,4,8,10-tetraen-7-yl Methanesulfonate (12). Freshly prepared 11 (2.63 g,

16.41 mmol), dissolved in CH₂Cl₂(90 ml) was cooled to 08. At this temp., Et₃N (9.2 ml, 66.0 mmol) and then

MsCl (2.55 ml, 32.81 mmol) were added. The dark yellow mixture was stirred during 1 h at r.t. To hydrolyze the

excess MsCl, ice-cold 0.25m HCl (100 ml) was added. The org. phase was washed (2Â) with sat. aq. NaHCO₃and

with H₂O. The soln. was filtered over cotton wool, and the solvent was distilled off. The dark brown residue was

dissolved in a 1 :1 mixture of hexane/t-BuOMe and filtered over alumina to give 12 (3.79 g, 97%). Yellow, very

unstable oil. R_f(hexane/TBME 2 :1) 0.34. IR (film): 3016m, 2936m, 1607w, 1447m, 1338vs, 1279w, 1175vs, 1129m,

1039m, 1010m, 966s, 920vs, 853s, 784s, 747s, 719s, 651w, 568m, 521m, 448w. ¹H-NMR: 6.28 (dd, ³J(8,9) 9.3, ⁴J

3

4

1.9, H C(8)); 6.10 (d, ³J(2,3) 11.9, H C(2)); 5.63 (ddd, ³J(4,5) 12.2, J(4,3) 6.5, J 1.8, H C(4)); 5.89

(dd, ³J(9,8) 9.3, ³J(9,10) 4.4, H C(9)); 5.63 ± 5.58 (m, H C(5,10)); 4.14 (dd, ABX, J_AB 14.6, J_AX 8.5,

2

3

2

H_endoC(6)); 3.81 (dd, A of AB, J_AB 11.9, ⁴J 1.9, CH₂(11)); 3.59 (d, B of AB, J_AB 11.9, CH₂(11)); 3.06

(s, MeOSO₂); 2.44 (ddd, B of ABX, J_AB 14.6, J_BX 7.8, ⁴J 2.0, H_exoC(6)). ¹³C-NMR: ca. 138 (s, C(1));

2

3

135.10, 128.57, 126.84, 126.11, 123.32 (7d, C(2) ± C(5), C(8) ± C(10)); ca. 77 (s, C(7)); 40.41 (q, MeOSO₂); 35.78,

.

33.00 (2t, C(6,11)). GC/MS (C₁₂H₁₄O₃S; 238.31): 238 (83, M ), 159 (89, [M MeSO₂] ), 142 (34, [M

MsOH] ), 141 (67), 131 (100), 115 (88), 103 (21), 91 (86), 77 (33).

.

1.1.6.4. Elimination of MsOH from 12. To a soln. of freshly prepared 11 (3.37 g, 14.14 mmol) in Et₂O

(100 ml) was added 1,5-diazabicyclo[4.3.0]non-5-ene (DBN; 8.5 ml, 71.2 mmol) at r.t. After stirring during 1 h,

the initially yellow soln. turned orange, and the formation of a brown, polymer-like precipitate was observed.

Stirring was continued during 15 h at r.t. After addition of a second amount of DBN (4.0 ml, 33.5 mmol), stirring

was continued for 2 additional h. Then, the now dark orange-colored mixture was diluted with pentane (200 ml),

poured on ice, and washed several times with sat. aq. NaCl soln. After drying (Na₂SO₄) and careful removal of

the solvent mixture by distillation, the brown-to-orange-colored residue was subjected to CC (alumina;

pentane) to yield 2 (1.07 g, 53%). Orange-colored, volatile and unstable oil. R_f(pentane) 0.70. UV/VIS

(hexane): l_max257 (sh, 4.12), 281 (4.54), 297 (sh, 3.96), 326 (3.65), 485 (2.63); l_min228 (3.79), 315 (3.63). IR

(film): 3015s, 2920m, 2864m, 1508w, 1456s, 1399w, 1369w, 1306w, 1248m, 1223w, 1078vw, 1031w, 984m, 897s,

858m, 839m, 759vs, 685s, 656vs, 570m. ¹H-NMR (600 MHz, (D₆)acetone; see Fig. 2): 8.13 (d, ³J(2,3) ³J(6,5)

9.0, H C(2,6)); 7.55 (d, ³J(8,9) ³J(8,10) 7.1, H C(8,10)); 7.38 (td, ³J(4,3) ³J(4,5) 11.1, ⁴J 0.9, H C(4));

7.31 (dd, ³J(3,4) ³J(5,4) 11.0, ³J(3,2) ³J(5,6) 8.3, H C(3,5)); 6.88 (t, ³J(9,8) ³J(9,10) 7.2, H C(9));

0.75 (d, ²J(11_b,11_a) 9.8, H_bC(11)); 1.22 (dt, ^{2 J}(11_a, 11_b) 9.8, ⁴J(11_a,8) ⁴J(11_a, 10) 1.6, H_aC(11).

9

)

AM1 Calculations of the lowest-energy conformation of 10 show H_endoC(6) in a more or less co-linear

arrangement with H C(5) (V(H C(5) C(6)

H

_endo) 188), i.e., it lies almost in the p-plane of the

adjacent CC bond, whereas H_exoC(6) extends into the p-cloud of the tetraene system

(V(H C(5) C(6) H_exo) 1368). These structural features explain well the large chemical-shift

difference between H_endoand H_exo

.

Helvetica Chimica Acta ± Vol. 84 (2001)

1041

¹³C-NMR (150 MHz, (D₆)acetone): 160.64 (s, C(1,7)); 143.53 (d, C(2,6)); 132.97 (d, C(8,10)); 129.77 (d, C(9));

.

128.35 (d, C(4)); 124.89 (d, C(3,5)); 34.63 (t, C(11)). GC/MS (C₁₁H₁₀; 142.20): 142 (75, M ), 141 (99), 139 (66),

126 (10), 115 (100, [C₉H₇( inden-1-yl)] ), 102 (19), 98 (7), 91 (15), 89 (65), 87 (26), 77 (22), 74 (32), 70 (26),

65 (32), 63 (68), 57 (30), 51 (57).

1.2. Bicyclo[5.3.1]undeca-1,3,5,7,9-pentaene-8-carbaldehyde ( Homoazulene-8-carbaldehyde; 27). The

Vilsmeyer reagent was prepared from DMF (4 ml) and POCl₃(0.825 ml, 9.00 mmol) at 08. It was added

dropwise with a syringe to a vigorously stirred soln. of 2 (1.067 g, 7.50 mmol) in DMF (6 ml). A change of the

color of the mixture from orange to dark red was observed. After 15 min, ice-cold H₂O (20 ml) was added, and

the mixture was neutralized with 4n aq. NaOH. The aq. phase was extracted (3Â) with CH₂Cl₂, and the

combined org. layers were washed (2Â) with 4n NaOH and once with sat. aq. NaHCO₃soln. After filtration

over cotton wool, the solvent was distilled off, and the residue, which still contained DMF, was freed from the

latter in h.v. at 458. CC (silica gel; hexane/t-BuOMe 3 :1) provided pure 27 (0.856 g, 67%). Red, unstable oil. R_f

(hexane/TBME 3 :1) 0.34. UV/VIS (MeCN): l_max218 (4.10), 238 (4.12), 293 (4.25), 378 (3.82); l_min202 (3.98),

225 (4.08), 257 (3.96), 341 (3.59). IR (film): 3020w, 2818w, 2719w, 1659vs, 1566w, 1473s, 1427m, 1347w, 1227s,

1207s, 1075w, 980m, 905w, 873w, 804s, 762m, 721w, 659s, 616m, 589w, 551w, 525vw. ¹H-NMR (600 MHz, CDCl₃):

10.05 (s, CHO)); 8.66 (d, ³J(6,5) 8.9, H C(6)); 8.12 (d, ³J(2,3) 8.6, H C(2)); 7.55 (d, ³J(9,10) 7.2,

H

C(9)); 7.53 (dd, ³J(10.4, ³J 8.9, H C(5)); 7.49 ± 7.44 (m, H C(3,4,10)); 0.03 (d, ²J(11_b, 11_a) 9.9,

H_bC(11)); 1.00 (dd, ²J(11_a, 11_b) 9.9, ⁴J(11_a, 10) 1.6, H_aC(11). ¹³C-NMR (150 MHz, CDCl₃): 187.10

(d, CHO)); 158.65 (s, C(1)); 154.51 (s, C(7)); 143.72 (d, C(6)); 143.20 (d, C(2)); 141.18 (s, C(8)); 136.30

(d, C(9)); 130.16 (d, C(10)); 128.68 (d, C(4)); 128.10 (d, C(5)); 126.60 (d, C(3)); 34.84 (t, C(11)). GC/MS

.

(C₁₂H₁₀O; 170.21): 170 (32, M ), 169 (27, [M 1] ), 141 (100, [M CHO] ), 139 (34), 126 (3), 115 (66), 102

(4), 91 (4), 89 (22), 77 (6), 63 (18).

The formation of azulene under the same conditions as described above gave azulene-1-carbaldehyde as a

violet oil in a yield of 96%.

1.3. 8-[(E)- and (Z)-2-Phenylethenyl]bicyclo[5.3.1]undeca-1,3,5,7,9-pentaene ( 8-[(Z)- and (E)-2-Phenyl-

ethenylhomoazulene; (Z)-28 and (E)-28). To a suspension of benzyl(triphenyl)phosphonium chloride (2.29 g,

5.89 mmol) in THF (11 ml) was added at r.t. t-BuOK (0.70 g, 6.24 mmol) in THF (22 ml). The red ylide soln.

prepared was stirred during 1 h at r.t. and then cooled to 788. A soln. of 27 (0.310 g, 1.82 mmol) in THF

(10 ml) was added, and stirring was continued during 15 min at 788. The Wittig reaction was completed by

stirring at r.t. during 1 h. H₂O (10 ml) and Et₂O (ca. 50 ml) were added, and the org. layer was washed with sat.

aq. NaHCO₃soln. and then dried (Na₂SO₄). The solvent mixture was concentrated until Ph₃PO had

precipitated. It was filtered and washed with Et₂O. The filtrate was concentrated again, the residue taken up in

hexane and filtered over a short alumina column to give a 1.6 :1 mixture of (Z)-28 and (E)-28 (0.440 g, 99%) as a

red unstable oil. For the separation of the (Z)- and (E)-isomers, the mixture was carefully chromatographed on

alumina (400 g) with hexane to give, as a first fraction, pure (Z)-28 (0.093 g), followed by a larger mixed fraction

of (Z)-28 and (E)-28. Finally, pure (E)-28 (0.075 g) was obtained as a third fraction.

Data of (Z)-28: Dark red oil. R_f(hexane) 0.31. UV/VIS (hexane): l_max224 (4.27), 264 (4.38), 304 (4.35),

384 (3.95), 484 (sh, 2.79); l_min215 (4.24), 243 (4.20), 277 (4.14), 343 (3.63). IR (film): 3019s, 2862w, 1598m,

1572vw, 1492s, 1446s, 1250w, 1179vw, 1155vw, 1074w, 1028w, 961w, 918m, 872w, 820m, 774vs, 695vs, 663vs, 630w,

614w, 544m, 479w. ¹H-NMR (600 MHz, (D₆)acetone): 8.13 (d, ³J(2,3) 8.2, H C(2)); 7.73 (d, ³J(6,5) 8.4,

H

C(6)); 7.50 (d, ³J(7.5, H_oof Ph); 7.28 ± 7.26 (m, H_mof Ph, H C(10)); 7.20 ± 7.15 (m, H_pof Ph, H C(3,4));

7.02 ± 6.99 (m, H C(5)); 6.77 (d, A of AB, ²J_AB 12.0, H C(1')); 6.74 (d, ³J(9,10) 7.1, H C(9)); 6.51 (d, B of

AB, J_AB 12.0, H C(2')); 0.19 (d, ²J(11_b, 11_a) 9.8, H_bC(11)); 0.89 (dd, ²J(11_a, 11_b) 9.8, J(11_a, 10)

1.4, H_aC(11). ¹³C-NMR (150 MHz, (D₆)acetone): 157.58 (s, C(1)); 156.48 (s, C(7)); 144.02 (d, C(2)); 143.24

(s, C(8)); 140.64 (d, C(6)); 139.22 (s, C_ipsoof Ph); 131.46 (d, C(10)); 130.59 (d, C(9)); 129.74 (d, C_oof Ph); 129.30

(d, C(2')); 129.10 (d, C_mof Ph); 128.26 (d, C(4)); 127.92 (d, C_pof Ph); 126.99 (d, C(1')); 125.27 (d, C(3)); 125.16

(d, C(5)); 34.67 (t, C(11)). GC/MS (C₁₉H₁₆; 244.34): 244 (90, M ), 243 (89, [M H] ), 228 (90), 215 (88), 202

(87), 189 (61), 165 (98), 152 (79), 141 (87), 139 (76), 128 (65), 121 (52), 115 (100), 114 (82), 107 (48), 102 (52),

91 (81), 77 (72).

2

4

.

Data of (E)-28: Red oil. R_f(hexane) 0.18. UV/VIS (hexane): l_max228 (4.25), 263 (4.25), 314 (4.36), 399

(4.24); l_min214 (4.13), 245 (4.12), 276 (4.06), 343 (3.72). IR (film): 3023s, 2955m, 2923m, 2866m, 1796vw, 1594s,

1495s, 1447s, 1251m, 1155vw, 1072w, 1028w, 958vs, 897m, 814s, 796m, 777vs, 717vs, 692vs, 658vs, 616s, 548s, 496w.

¹H-NMR (500 MHz, (D₆)acetone): 8.21 (d, ³J(2,3) 9.6, H C(2)); 7.98 (d, ³J(6,5) 7.1, H C(6)); 7.67

(d, ³J(7.5, H_oof Ph); 7.63 (d, ³J(1',2') 16.4, H C(1')); 7.40 (t, ³J 7.7, H_mof Ph); 7.31 (d, ³J(2',1') 16.4,

H

C(2')); 7.31 ± 7.25 (m, H_pof Ph, H C(10)); 7.21 ± 7.17 (m, H C(4), H C(5)); 7.11 (t, ³J 9.7, H C(3));

4

6.92 (d, ³J(9,10) 6.6, H C(9)); 0.03 (d, ²J(11_b, 11_a) 9.8, H_bC(11)); 0.25 (dd, ²J(11_a, 11_b) 9.8, J(11_a,

1042

Helvetica Chimica Acta ± Vol. 84 (2001)

10) 1.2, H_aC(11)). ¹³C-NMR (125 MHz, (D₆)acetone): 156.62 (s, C(1)); 154.93 (s, C(7)); 146.08 (s, C(8));

145.01 (d, C(2)); 140.11 (d, C(6)); 138.63 (s, C_ipsoof Ph); 130.54 (d, C(10)); 129.51 (d, C_mof Ph); 128.90

(d, C(2')); 128.79 (d, C(4)); 128.13 (d, C_pof Ph); 127.18 (d, C_oof Ph); 126.41, 126.37 (2d, C(1') and C(9)); 125.48

.

(d, C(5)); 124.86 (d, C(3)); 35.15 (t, C(11)). GC/MS (C₁₉H₁₆; 244.34): 244 (67, M ), 243 (60, [M H] ), 228

(78), 226 (77), 215 (88), 202 (81), 189 (53), 178 (7), 165 (85), 152 (70), 141 (87), 139 (77), 128 (55), 121 (36),

115 (100), 114 (77), 102 (47), 91 (80), 77 (66).

2. Homoheptalene-dicarboxylates. ± 2.1. Dimethyl Bicyclo[5.5.1]trideca-1,3,5,7,9,11-hexaene-2,3- and -4,5-

dicarboxylate ( Dimethyl Homoheptalene-2,3- and -4,5-dicarboxylate; 13 and 3, resp.). Homoazulene (2;

0.117 g, 0.823 mmol) was dissolved in ClCH₂CH₂Cl (0.5 ml) under Ar in a Schlenk vessel. ADM (0.30 ml,

2.44 mmol) was added to the orange-colored soln., which was heated under Ar during 30 h at 828. ClCH₂CH₂Cl

and the excess of ADM was removed from the now brown-red mixture at 508 in h.v. The residue was subjected to

CC (silica gel; hexane/t-BuOMe 2 :1) to give a 4 :1 mixture of 3 and 13 (0.225 g, 96%) as a blood-red oil. For the

separation of 3 and 13, the mixture was again carefully chromatographed on silica gel (400 g) with hexane/t-

BuOMe 4 :1. This led to a fraction of strongly enriched 3, from which pure 3 crystallized (0.090 g) after it had

been dissolved in hexane/t-BuOMe 10 :1. The combined residues (from CC and the mother liquor of

crystallization) were once more chromatographed on silica gel (400 g) with hexane/t-BuOMe 4 :1. Finally, from

fractions enriched in 13, the latter homoheptalene was obtained in pure crystalline form (9 mg) by

crystallization from hexane/t-BuOMe 15 :1.

Data of 3: Red-orange prisms. M.p. 139 ± 1408 (hexane/t-BuOMe) ([2]: 1398). R_f(hexane/TBME 2 :1)

0.180. UV/VIS (hexane): l_max270 (4.56), 330 (sh, 3.65), 419 (v. br.; 2.67); l_min212 (4.03), 392 (2.63). IR (KBr):

3006w, 2950w, 1725vs, 1700vs, 1610s, 1435s, 1377w, 1301m, 1275vs, 1243vs, 1069vs, 1045s, 994m, 954w, 934m,

885m, 865m, 854m, 789s, 764w, 753m, 723s, 699w, 669w, 653w, 560w, 522w, 482vw. ¹H-NMR (600 MHz, CDCl₃;

see Fig. 4,a, and Table 2): 7.19 (s, H C(6)); 6.99 (d, ³J(3,2) 4., H C(3)); 6.11 (d, ³J(8,9) 4.7, H C(8)); 6.09

(d, ³J(12,11) 12.3,

H

C(12)); 5.92 (d, ³J(2,3) 4.3,

H

C(2)); 5.67 (dd, ³J(9,10) 11.5, ³J(9,8) 4.8,

H

C(9)); 5.61 (dd, ³J 6.6, H C(11)); 5.60 (dd, ³J(10,9) 11.5, ³J(10,11) 5.5, H C(10)); 5.07 (d, ²J_AB

11.3, A of AB, H_synC(13)); 4.97 (d, ²J_AB 11.3, B of AB, H_antiC(13)); 3.71 (s, MeOCO C(4)); 3.68

(s, MeOCO C(5)). ¹³C-NMR (150 MHz, CDCl₃): 168.45 (s, MeOCO C(5)); 167.55 (s, MeOCO C(4));

145.19 (s, C(1)); 144.35 (d, C(6)); 144.11 (d, C(3)); 143.87 (s, C(7)); 137.16 (d, C(8)); 135.97 (d, C(12)); 130.43

(d, C(9)); 130.36 (s, C(5)); 130.26 (s, C(4)); 129.49 (d, C(10)); 129.28 (d, C(11)); 125.99 (d, C(2)); 52.22

(q, MeOCO C(5)); 52.09 (q, MeOCO C(4)); 31.65 (t, C(13)). GC/MS (C₁₇H₁₆O₄; 284.32): 284 (96, M ), 259

(5), 253 (60, [M MeO] ), 252 (32, [M MeOH] ), 237 (22), 225 (87, [M COOMe] ), 224 (82), 209 (67),

.

193 (79), 181 (50), 166 (70, [M 2 COOMe] ), 165 (100), 164 (82), 163 (78), 153 (80), 152 (85), 139 (82), 127

(75), 115 (81), 102 (18), 91 (22), 89 (28),77 (32).

In addition, the structure of 3 was determined by an X-ray crystal-structure analysis (cf. Fig. 5,a and

Tables 3 and 8).

Data of 13: Red prisms. M.p. 108 ± 1128. R_f(hexane/t-BuOMe 2 :1) 0.182. UV/VIS (hexane): l_max279 (4.51),

420 (v. br.; 2.70); l_min222 (3.98), 382 (2.64). IR (KBr): 3417vw, 3007w, 2957m, 2925w, 2853w, 1719vs, 1703vs,

1627m, 1610w, 1597m, 1584w, 1564s, 1535w, 1447m, 1428s, 1369w, 1263vs, 1244vs, 1203vs, 1154s, 1118s, 1059s,

1039s, 1005m, 982m, 941m, 921w, 903m, 893w, 879m, 860m, 840w, 822m, 797m, 785s, 738m, 727m, 712w, 676m,

651m, 622w, 574w, 537vw, 519vw, 503vw, 457vw. ¹H-NMR (600 MHz, CDCl₃; see Fig. 4,b, and Table 3): 7.71

(d, ³J(12,11) 12.9, H C(12)); 6.93 (d, ³J(4,5) 6.4, H C(4)); 6.00 (ddd, ³J(11,12) 12.9, J(11,10) 7.6, J

0.9, H C(11)); 5.99 (d, ³J(8,9) 5.2, H C(8)); 5.99 (d, ³J(6,5) 12.8, H C(6)); 5.86 (dd, ³J(9,10) 12.2,

³J(9,8) 5.2, H C(9)); 5.72 (dd, ³J(10,9) 12.2, ³J(10,11) 5.2, H C(10)); 5.68 (dd, ³J(5,6) 12.8, ³J(5,4)

3

4

6.4, H C(5)); 5.41 (dd, A of ABX, ²J_AB 11.1, ⁴J_AX 1.2, H_synC(13)); 4.55 (dd, B of ABX, ²J_AB 11.1, ⁴J_BX

0.8, H_antiC(13)); 3.69 (s, MeOCO C(3)); 3.68 (s, MeOCO C(2)). ¹³C-NMR (150 MHz, CDCl₃): 167.20

(s, MeOCO C(3)); 166.98 (s, MeOCO C(2)); 154.55 (s, C(1)); 143.80 (s, C(7)); 140.76 (d, C(4)); 137.06

(d, C(6)); 133.98 (s, C(3)); 133.70 (d, C(8)); 132.56 (d, C(11)); 131.88 (d, C(9)); 130.23 (d, C(12)); 128.48

(d, C(10)); 128.08 (s, C(2)); 127.54 (d, C(5)); 52.18 (q, MeOCO C(2)); 51.51 (q, MeOCO C(3)); 34.78

.

(t, C(13)). GC/MS (C₁₇H₁₆O₄; 284.32): 284 (35, M ), 252 (9, [M MeOH] ), 237 (10), 225 (20, [M

.

COOMe] ), 209 (10), 193 (13), 181 (9), 166 (58, [M 2 COOMe] ), 165 (100), 164 (35), 163 (28), 152

(32), 139 (20), 128 (10), 115 (19), 77 (5).

In addition, the structure of 13 was determined by an X-ray crystal-structure analysis (cf. Fig. 5,b, and

Tables 3 and 8).

2.1.1. Dimethyl ()-(pR)- and ( )-(pS)-Homoheptalene-4,5-dicarboxylate ((pR)-3 and (pS)-3). Com-

pound (Æ)-3 of 2.1 (11.8 mg) was separated on a semi-prep. Chiralcel OD column (250 Â 20 mm) with hexane/

i-PrOH (20%) as eluant and a flow rate of 9 ml/min, which led to ()-(pR)-3 (4.7 mg) in the first fraction and

Helvetica Chimica Acta ± Vol. 84 (2001)

1043

(

)-(pS)-3 (5.2 mg) in the second one. Both enantiomers were obtained as a red oil. Attempts for crystallization

were not undertaken.

Data of ()-(pR)-3: t_R(conditions see above) 21.0 min. CD (hexane; r.t.; cf. Fig. 6): 425 (br., pos. max.;

6.04), 364 (pos. min.; 3.60), ca. 330 (pos. sh; 6.0), 296 (pos. max.; 7.97), 281 (0), 273 (neg. max.; 5.92), 264 (0),

258.5 (pos. max.; 4.03), 244 (pos. min.; 0.97), 227.5 (pos. max.; 10.98).

Data of ( )-(pS)-3: t_R(conditions see above) 56.0 min. CD (hexane; r.t.; cf. Fig. 6): 424 (br., neg. max.;

5.77), 364 (neg. min.; 3.58), ca. 325 (neg. sh; 6.7); 295.5 (neg. max.; 8.43), 281 (0), 272.5 (pos. max.; 5.18), 264

(0), 256 (neg. max.; 4.80), 244 (neg. min.; 1.89), 228 (neg. max.; 11.00).

2.2. Dimethyl 2-[(Z)- and (E)-2-Phenylethenyl]bicyclo[5.5.1]trideca-1,3,5,7,9,11-hexaene-4,5-dicarboxylate

( Dimethyl 2-[(Z)- and (E)-2-Phenylethenyl]homoheptalene-4,5-dicarboxylate; (Z)-29 and (E)-29). The 1.6 :1

mixture of (Z)-28 and (E)-28 (0.221 g, 0.904 mmol) was dissolved in ClCH₂CH₂Cl (2.0 ml) in a Schlenk vessel.

ADM (0.34 ml, 2.39 mmol) was added, and the stirred mixture was heated under Ar during 18 h at 828, whereby

the color of the mixture changed from red to brown. The solvent and the excess ADM were removed at 508 in

h.v. The residue was subjected to CC (silica gel; hexane/TBME 3 :1) to give a 5 :1 mixture of (Z)-29 and (E)-29

as a brown-red oil (0.147 g, 42%). For the separation of the (Z)- and (E)-isomer, the mixtures of three runs were

combined (0.332 g) and carefully chromatographed again (silica gel (400 g); hexane/t-BuOMe 4 :1). A first

fraction gave pure (Z)-29 as a brown-red oil (0.138 g). The following fractions delivered enriched (E)-29, which

crystallized after ca. 7 d as brown-red needles (0.034 g).

Data of (Z)-29: R_f(hexane/t-BuOMe 2 :1) 0.313. UV/VIS (hexane; see Fig. 9): l_max243 (sh, 4.40), 267

(4.54), 310 (sh, 4.12), 445 (2.77); l_min219 (4.28), 408 (2.77). IR (CHCl₃): 3025m, 3015m, 2952w, 1715vs, 1615w,

1493vw, 1436s, 1250vs, 1205w, 1128vw, 1085m, 1031w, 985vw, 937vw, 854vw, 829vw. ¹H-NMR (600 MHz, CDCl₃):

7.47 (d, ³J 7.5, H_oof Ph); 7.25 (s, H C(6)); 7.23 (t, ³J 7.6, H_mof Ph)); 7.12 (t, ³J 7.3, H_pof Ph)); 6.87

(s, H C(3)); 6.60 (d, ³J(2',1') 12.0, H C(2')); 6.35 (d, ³J(12,11) 13.0, H C(12)); 6.16 (d, ³J(8,9) 5.1,

H

C(8)); 6.10 (d, ³J(1',2') 12.0,

H

C(1')); 5.67 (dd, ³J(9,10) 12.4, ³J(9,8) 5.1,

H C(9)); 5.57

(dd, ³J(10,9) 12.4, ³J(10,11) 7.6, H C(10)); 5.51 (dd, ³J(11,12) 13.0, ³J(11,10) 7.6, H C(11)); 5.23

(d, ²J 11.6, H_synC(13)); 4.86 (d, ²J 11.6, H_antiC(13)); 3.67 (s, MeOCO C(4)); 3.64 (s, MeOCO C(5)).

¹³C-NMR (150 MHz, CDCl₃): 168.27 (s, MeOCO C(5)); 167.79 (s, MeOCO C(4)); 146.05 (d, C(3)); 144.65

(d, C(6)); 144.05 (s, C(7)); 139.10 (s, C(1)); 138.12 (d, C(8)); 136.98 (s, C_ipsoof Ph); 134.18 (s, C(2)); 133.33

(d, C(2')); 133.12 (d, C(12)); 130.78 (s, C(4)); 130.20 (d, C(9)); 129.70 (d, C(10)); 129.63 (s, C(5)); 129.36

(d, C(11)); 129.19 (d, C_oof Ph); 128.30 (d, C_mof Ph); 127.30 (d, C_pof Ph); 125.32 (d, C(1')); 52.36

(q, MeOCO C(4)); 52.27 (q, MeOCO C(5)); 32.84 (t, C(13)). GC/MS (C₂₅H₂₂O₄; 386.45): 386 (87, M ),

.

354 (7), 327 (36, [M COOMe] ), 326 (22), 311 (11), 295 (39), 294 (15), 268 (42, [M 2 COOMe] ), 267

(93), 266 (60), 265 (90), 252 (100), 239 (43), 229 (16), 227 (18), 207 (11), 202 (16), 189 (43), 178 (18), 165 (38),

152 (19), 139 (14), 133 (27), 126 (19), 115 (29), 103 (13), 91 (40), 77 (10).

Data of (E)-29: Brown-red needles. M.p. 188 ± 1898 (CH₂Cl₂/Et₂O). R_f(hexane/t-BuOMe 2 :1) 0.291. UV/

VIS (hexane; see Fig. 9): l_max235 (4.36), 262 (4.39), 318 (4.61),445 (2.91); l_min218 (4.21), 244 (4.34), 281 (4.33),

420 (2.91). IR (KBr): 3404vw, 3002w, 2947w, 1711vs, 1615w, 1495vw, 1435m, 1310w, 1255vs, 1159vw, 1079m,

1067m, 1038w, 968vw, 951w, 934w, 858vw, 784w, 754w, 723m, 693w, 659vw, 527vw. ¹H-NMR (600 MHz, CDCl₃):

7.45 (d, ³J 7.5, H_oof Ph); 7.355 (d, ³J(1',2') 15.9, H C(1')); 7.346 (s, H C(3)); 7.32 (t, ³J 7.7, H_mof Ph)); 7.23

(t, ³J 7.4, H_pof Ph)); 7.20 (s, H C(6)); 6.86 (d, ³J(12,11) 13.2, H C(12)); 6.61 (d, ³J(2',1') 15.9, H C(2'));

6.22 (d, ³J(8,9) 5.0, H C(8)); 5.85 ± 5.81 (m, H C(9), H C(11)); 5.77 (dd, ³J(10,9) 12.5, ³J(10,11) 7.4,

H

C(10)); 5.04 (d, ²J 11.7, H_synC(13)); 4.65 (d, ²J 11.7, H_antiC(13)); 3.79 (s, MeOCO C(4)); 3.71

(s, MeOCO C(5)). ¹³C-NMR (150 MHz, CDCl₃): 167.87 (s, MeOCO C(5)); 167.31 (s, MeOCO C(4));

145.30 (s, C(7)); 144.48 (d, C(3)); 143.67 (d, C(6)); 138.96 (s, C(1)); 138.27 (d, C(8)); 137.36 (s, C_ipsoof Ph);

134.11 (s, C(2)); 131.53 (s, C(5)); 131.10 (d, C(12)); 130.20 (d, C(11)); 129.79 (s, C(4)); 129.51 (d, C(10)); 129.34

(d, C(2')); 128.80 (d, C(9)); 128.61 (d, C_mof Ph); 127.70 (d, C_pof Ph); 126.63 (d, C_oof Ph); 123.70 (d, C(1'));

52.27 (q, MeOCO C(4)); 52.25 (q, MeOCO C(5)); 33.31 (t, C(13)). GC/MS (C₂₅H₂₂O₄; 386.45): 386 (98,

.

M

), 354 (8), 327 (44, [M COOMe] ), 326 (24), 311 (13), 295 (46), 294 (16), 268 (46, [M 2 COOMe] ),

267 (99), 266 (63), 265 (96), 252 (100), 239 (45), 229 (17), 227 (18), 215 (15), 202 (16), 189 (47), 178 (19), 165

(40), 152 (18), 133 (27), 115 (29), 103 (13), 91 (42), 77 (15). Anal. calc. for C₂₅H₂₂O₄(386.45): C 77.70, H 5.74;

found: C 77.48, H 5.75.

2.3. 2-[(Z)- and (E)-2-Phenylethenyl]bicyclo[5.5.1]trideca-1,3,5,7,9,11-hexaene-4,5-dimethanol (2-[(Z)-

and (E)-2-Phenylethenyl]homoheptalene-4,5-dimethanol; (Z)-31 and (E)-31, resp.) and 2-[(Z)-2-Phenyl-

ethenyl]-6-oxatricyclo[8.5.1.0^4,8]hexadeca-2,4(8),9,11,13,15-hexaen-5-one (1,3-Dihydro-10-[(Z)-2-phenyl-

ethenyl]homoheptaleno[4,3-c]furan-1-one; (Z)-32). Into a soln. of (Z)-29 (0.051 g, 0.132 mmol) in toluene

(15 ml) was injected 1.5m DIBAH in toluene (0.105 ml, 0.158 mmol) at 908. TLC Control after 3 h revealed

1044

Helvetica Chimica Acta ± Vol. 84 (2001)

still the presence of a large amount of starting material. Therefore, additional DIBAH soln. (ca. 0.05 ml,

0.075 mmol) was injected. After a further reduction time of 1 h, all (Z)-29 had reacted, and the soln. was

allowed to warm to r.t. Ice-H₂O (10 ml) was added. The toluene phase was washed with 2n aq. HCl soln., H₂O,

and sat. NaCl soln., and then dried (Na₂SO₄). CC (silica gel; hexane/t-BuOMe 2 :1) of the residue of the toluene

soln. gave, in a first fraction, (Z)-31 (0.012 g, 28%) as a brown-yellow oil, which, on standing for 1 d in the ice-

box, isomerized quantitatively into (E)-31, and in a second fraction (Z)-32 (0.008 g, 19%) as a violet colored oil,

which showed no tendency to undergo (Z)/(E)-isomerization.

Data of (Z)-31: R_f(hexane/t-BuOMe 2 :1) 0.10. ¹H-NMR (CDCl₃): 7.51 (d, ³J 7.2, H_oof Ph); 7.28 (t, ³J

7.1, 2 H_mof Ph)); 7.20 (t, ³J 7.2, H_pof Ph)); 6.51, 6.22 (2d, AB, ³J_AB 12.0, H C(1',2')); 6.38 (d, ³J(12,11) 12.2,

H

C(12)); 6.10 (s, H C(3)); 5.82 (d, ³J(8,9) 4.9, H C(8)); 5.59 (dd, ³J(9,10) 11.5, ³J(9,8) 5.1, H C(9));

5.53 ± 5.48 (m, H C(6,10,11), 1 H of CH₂(13)); 5.01 (d, ²J 11.1, 1 H of CH₂(13)); 4.10 (d, A of AB(1), ²J_AB

12.5, CH₂C(4 or 5); 4.08 (dd, A of AB(2), ²J_AB 12.5, J 1.1, CH₂C(5 or 4)); 3.93 (d, B of AB(2), ²J_AB 12.5,

CH₂C(4 or 5)); 3.90 (d, B of AB(2), ²J_AB 12.5, CH₂C(5 or 4)); 1.19 (br. s, 2 OH)).

Data of (E)-31: Brown-yellow oil. R_f(hexane/t-BuOMe 2 :1) 0.10. UV/VIS (hexane; see Fig. 10): l_max251

(4.34), 317 (4.43), 326 (sh, 4.39), 362 (sh, 3.64), 430 (sh, 3.20); l_min220 (4.13), 280 (4.02); tailing up to >550. IR

(film): 3350m, 3005m, 2956m, 2928m, 2882w, 1597w, 1556vw, 1492w, 1447m, 1375w, 1303w, 1265m, 1155vw,

1128w, 1103w, 1055w, 1014s, 953m, 908vs, 883m, 844w, 783w, 734vs, 695s, 668w, 648m, 552w, 519w. ¹H-NMR

(600 MHz, CDCl₃): 7.42 (d, ³J 7.8, H_oof Ph); 7.40 (d, ³J(1',2') 16.2, H C(1')); 7.30 (t, ³J 7.7, H_mof Ph)); 7.20

(t, ³J 7.3, H_pof Ph)); 6.81 (d, ³J(12,11) 13.1,

H

C(12)); 6.52 (d, ³J(2',1') 15.9,

H C(2')); 6.08

(s, H C(3,6)); 5.87 (d, ³J(8,9) 5.1,

H

C(8)); 5.78 ± 5.74 (m, H C(9,1)); 5.67 (dd, ³J(10,9) 12.5,

³J(10,11) 7.5, H C(10)); 5.30, 4.64 (2d, ²J 11.2, CH₂(13)); 4.47 (dd, A of AB(1), J_AB 12.5, J 1.1,

2

CH₂C(4)); 4.31 (d, B of AB(1), J_AB 12.5, CH₂C(4)); 4.31, 4.25 (d, AB(2), J_AB 12.4, CH₂C(5)); 2.09

(br. s, 2 OH)). ¹³C-NMR (150 MHz, CDCl₃): 147.36 (s, C(7)); 141.58, 140.79 (2s, C(4,5)); 140.22 (s, C(1)); 137.67

(s, C_ipsoof Ph); 135.24 (d, C(6)); 135.01 (s, C(2)); 133.16 (d, C(3)); 133.22 (d, C(8)); 130.56 (d, C(9)); 130.41

(d, C(12)); 128.59 (d, C_mof Ph); 128.24 (d, C(10)); 127.93 (d, C(11)); 127.54 (d, C(2')); 127.37 (d, C_pof Ph);

126.46 (d, C_oof Ph); 125.08 (d, C(1')); 68.77 (t, HOCH₂C(5)); 66.48 (t, HOCH₂C(4)); 33.78 (t, C(13)).

Data of (Z)-32: R_f(hexane/t-BuOMe 2 :1) 0.49. UV/VIS (hexane): l_max261 (4.54); l_min220 (4.20). IR

(CHCl₃): 3026w, 3016w, 1747s, 1596w, 1449w, 1301w, 1226w, 1135w, 1047w, 849w, 793vs, 767w, 760vw, 722m, 696w,

672m, 667w, 463vw, 456w. ¹H-NMR (600 MHz, CDCl₃): 7.62 (d, ³J 7.3, H_oof Ph); 7.34 (t, ³J 7.8, H_mof Ph));

7.23 (tt, ³J 7.4, ⁴J 1.1, H_pof Ph)); 6.68, 6.31 (2d, AB, J_AB 11.5, CH₂C(16)); 6.47 (d, ³J(2',1') 11.9,

2

3

H

C(2')); 5.99 (d, ³J(15,14) 4.9, H C(15)); 5.74 (d, ³J(1',2') ꢁ J(11,12) ꢁ 13, H C(11,1')); 5.64 (s, H C(3));

5.54 (s, H C(9)); 5.47 (dd, ³J(12,11) 13.0, ³J(12,13) 6.7, H C(12)); 5.25 (dd, ³J(14,13) 12.7, ³J(14,15)

4.9, H C(14)); 5.22 (dd, ³J(13,14) 12.7, ³J(13,12) 6.7, H C(13)); 4.43, 4.31 (2d, AB, J_AB 17.6, CH₂(7)).

2

¹³C-NMR (150 MHz, CDCl₃): 173.47 (s, OC(5); 157.13 (s, C(8)); 152.53 (s, C(10)); 145.60 (s, C(2)); 143.73

(s, C(1)); 136.53 (s, C_ipsoof Ph); 134.93 (d, C(14)); 134.25 (d, C(11)); 134.08 (d, C(12)); 132.79 (d, C(15)); 132.53

(d, C(2')); 131.43 (d, C(1')); 128.84 (d, C_oof Ph); 128.49 (d, C(13)); 128.23 (d, C_mof Ph); 127.35 (d, C_pof Ph);

123.39 (s, C(4)); 122.24 (d, C(3)); 121.75 (d, C(9)); 70.33 (t, C(7)); 36.12 (t, C(16)).

3. X-Ray Crystal-Structure Determinations¹⁰). ± 3.1 Diacetoxy 9. A crystal, obtained from pentane, was

mounted on a glass fiber and used for a low-temperature X-ray structure determination. All measurements were

made on a Rigaku AFC5R diffractometer using graphite-monochromated MoK_aradiation (l 0.71069 ) and

a 12-kW rotating anode generator. The unit-cell constants and an orientation matrix for data collection were

obtained from a least-squares refinement of the setting angles of 25 carefully centered reflections in the range

388 < 2q < 408. The w/2q scan mode was employed for data collection, where the w scan width was (1.21 0.35 ´

tan q)8 and the w scan speed was 128 min ¹. The weaker reflections [I < 10s(I)] were rescanned up to a

maximum of 4 scans and the counts were accumulated. Stationary background counts were recorded on each

side of the reflection with a peak/background counting-time ratio of 2 :1.

The intensities of three standard reflections were measured after every 150 reflections and remained stable

throughout the data collection. The intensities were corrected for Lorentz and polarization effects. Azimuthal

scans of several reflections indicated no need for an absorption correction. The space group was uniquely

10

)

Crystallographic data (excluding structure factors) for the structures reported in this paper have been

deposited with the Cambridge Crystallographic Data Center as supplementary publication Nos. CCDC-

159574, 159575, and 159576 for 3, 9, and 13, resp. Copies of the data can be obtained, free of charge, on

application to the CCDC, 12 Union Road, Cambridge CB21EZ, UK (fax: 44-(0)1223-336033; email:

deposit@ccdc.cam.ac.uk).

Helvetica Chimica Acta ± Vol. 84 (2001)

1045

determined by the systematic absences. Equivalent reflections were merged. Data collection and refinement

parameters are given in Table 8.

The structure was solved by direct methods with SHELXS97 [26], which revealed the positions of all non-

H-atoms. The non-H-atoms were refined anisotropically. All of the H-atoms were located in a difference

electron-density map, and their positions were allowed to refine together with individual isotropic displacement

parameters. Refinement of the structure was carried out on F by means full-matrix least-squares procedures,

which minimized the function Sw(j F_oj j F_cj ) [26]. The weighting scheme was based on counting statistics and

included a factor to down-weight the intense reflections. Plots of Sw(j F_oj j F_cj )²vs. j F_oj , reflection order in

data collection, sin q/l, and various classes of indices showed no unusual trends. A correction for secondary

extinction was applied.

Neutral-atom-scattering factors for non-H-atoms were taken from [27a], and the scattering factors for H-

atoms were taken from [28]. Anomalous dispersion effects were included in F_c[29]; the values for f ' and f '' were

Table 8. Crystallographic Data of Compounds 9, 3, and 13

9

3

13

Crystallized from

pentane

C₁₅H₁₈O₄

262.30

colorless, prism

0.22 Â 0.40 Â 0.50

173 (1)

t-BuOMe/hexane

C₁₇H₁₆O₄

284.31

red-orange, prism

0.27 Â 0.32 Â 0.38

173 (1)

t-BuOMe/hexane

C₁₇H₁₆O₄

284.31

red, prism

0.25 Â 0.30 Â 0.50

173 (1)

Empirical formula

Formula weight [g mol

Crystal color, habit

Crystal dimensions [mm]

Temp. [K]

1

]

Crystal system

Space group

monoclinic

P2₁/c

triclinic

P1

monoclinic

P2₁/n

Å

Z

4

2

4

Reflections for cell determination

25

2q range for cell determination [8]

38 ± 40

39 ± 40

34 ± 39

Unit-cell parameters

a []

b []

c []

a [8]

b [8]

g [8]

V [³]

F(000)

D_x[g cm

m(MoK_a) [mm

Scan type

8.554 (2)

12.417 (2)

12.928 (1)

90

95.67 (1)

90

1366.4 (4)

560

1.275

9.930 (2)

10.055 (1)

7.504 (1)

94.88 (1)

109.84 (1)

82.52 (1)

698.1 (2)

300

11.891 (2)

6.997 (4)

17.072 (2)

90

90.527 (9)

90

1420.3 (8)

600

1.329

3

]

1.352

0.0959

w/2q

1

]

0.0917

w/2q

0.0943

w/2q

2q_(max)[8]

55

Total reflections measured

Symmetry independent reflections

R_int

Reflections used [I > s(I)]

Parameters refined

Reflection/parameter ratio

Final R

3490

3129

0.015

2514

245

3390

3206

0.017

2466

255

9.67

0.0388

0.0342

0.005

3682

3248

0.021

2302

255

9.03

0.0436

0.0392

0.005

10.3

0.0397

0.0379

0.005

2.009

wR

Weights: p in w [s²(F_o) (pF_o)²]

1

Goodness of fit

1.858

1.726

6

7

Secondary extinction coefficient

1.8 (1) Â 10

2.6 (3)) Â 10

0.0004

0.23; 0.16

0.002 ± 0.003

2.9 (9)) Â 10

Final D_max/s

D1 (max; min) [e

s(d(C C)) []

0.0006

3

]

0.23; 0.19

0.002

0.22; 0.17

0.002 ± 0.003

1046

Helvetica Chimica Acta ± Vol. 84 (2001)

those from [27b]. The values of the mass attenuation coefficients are those from [27c]. All calculations were

performed with the teXsan crystallographic software package [30].

3.2. Homoheptalene-4,5-dicarboxylate 3. A crystal, obtained from TBME/hexane, was mounted on a glass

fiber and used for a low-temperature X-ray structure determination. All measurements were made on a Rigaku

AFC5R diffractometer using graphite-monochromated MoK_aradiation (l 0.71069 ) and a 12-kW rotating-

anode generator. The unit-cell constants and an orientation matrix for data collection were obtained from a

least-squares refinement of the setting angles of 25 carefully centered reflections in the range 398 < 2q < 408.

The w/2q scan mode was employed for data collection, where the w scan width was (1.52 0.35 ´ tan q)8 and the

w scan speed was 168 min ¹. The weaker reflections [I < 10s (I)] were rescanned up to a maximum of 4 scans

and the counts were accumulated. Stationary background counts were recorded on each side of the reflection

with a peak/background counting-time ratio of 2 :1.

The intensities of three standard reflections were measured after every 150 reflections and remained stable

throughout the data collection. The intensities were corrected for Lorentz and polarization effects. Azimuthal

scans of several reflections indicated no need for an absorption correction. The space group was determined

from packing considerations, a statistical analysis of intensity distribution, and the successful solution and

refinement of the structure. Equivalent reflections were merged. Data collection and refinement parameters are

given in Table 8. The structure was solved by direct methods using SIR92 [31], which revealed the positions of all

non-H-atoms. All further procedures were as described in Sect. 3.1. Aview of the molecule is shown in the Fig. 5.

The two seven-membered rings of 3 have quite localized CC bonds and adopt a W-shaped conformation.

3.3. Homoheptalene-2,3-dicarboxylate 13. A crystal obtained from t-BuOMe/hexane was mounted on a

glass fiber and used for a low-temperature X-ray structure determination. All measurements were made on a

Rigaku AFC5R diffractometer using graphite-monochromated MoK_aradiation (l 0.71069 ) and a 12-kW

rotating anode generator. The unit cell constants and an orientation matrix for data collection were obtained

from a least-squares refinement of the setting angles of 25 carefully centered reflections in the range 348 < 2q <

398. The w/2q scan mode was emplyed for data collection, where the w scan width was (1.42 0.35 ´ tan q)8, and

the w scan speed was 168 min ¹. The weaker reflections [I < 10s(I)] were rescanned up to a maximum of 4 scans

and the counts were accumulated. Stationary background counts were recorded on each side of the reflection

with a peak/background counting-time ratio of 2 :1.

The intensities of three standard reflections were measured after every 150 reflections and remained stable

throughout the data collection. The intensities were corrected for Lorentz and polarization effects. Azimuthal

scans of several reflections indicated no need for an absorption correction. The space group was uniquely

determined by the systematic absences. Equivalent reflections were merged. Data collection and refinement

parameters are given in Table 8. The structure was solved as described in Sect. 3.2. A view of the molecule is

shown in the Fig. 6. The two seven-membered rings of 13 have quite localized CC bonds and adopt again a W-

shaped conformation.

REFERENCES

[1] E. Vogel, H. Königshofen, J. Wassen, K. Müllen, J. F. M. Oth, Angew. Chem. 1974, 86, 229; Angew. Chem.,

Int. Ed. 1974, 13, 281.

[2] E. Vogel, H. Königshofen, J. Wassen, K. Müllen, J. F. M. Oth, Angew. Chem. 1974, 86, 777; Angew. Chem.,

Int. Ed. 1974, 13, 732.

[3] A. Mugnoli, M. Simonetta, J. Chem. Soc., Perkin Trans. 2 1976, 822.

[4] R. Destro, E. Ortoleva, M. Simonetta, R. Todeschini, J. Chem. Soc., Perkin Trans. 2 1983, 1227.

[5] L. T. Scott, M. A. Kirms, J. Am. Chem. Soc. 1982, 104, 3530.

[6] L. T. Scott, W. R. Brunsvold, M. A. Kirms, I. Erden, J. Am. Chem. Soc. 1981, 103, 5216.

[7] L. T. Scott, M. A. Kirms, H. Günther, H. von Puttkamer, J. Amer. Chem. Soc. 1983, 105, 1372.

[8] A. A. Briquet, P. Uebelhart, H.-J. Hansen, Helv. Chim. Acta 1996, 79, 2282.

[9] a) S. El Houar, H.-J. Hansen, Helv. Chim. Acta 1997, 80, 253; S. Maillefer-El Houar, P. Uebelhart, H.-J.

Hansen, Helv. Chim. Acta, in preparation; b) P. Ott, H.-J. Hansen, Helv. Chim. Acta 2001, 84, in

preparation.

[10] L. T. Scott, C. A. Sumpter, Org. Synth. 1993, Coll. Vol. VIII, 196.

[11] E. J. Corey, M. J. Chaykovsky, J. Am. Chem. Soc. 1965, 87, 1353.

[12] a) S. Moon, J. Org. Chem. 1964, 29, 3456; b) R. Munday, I. O. Sutherland, J. Chem. Soc., Chem. Commun.

1967, 569; c) R. J. Quellette, D. Miller, A. South Jr., R. D. Robins, J. Am. Chem. Soc. 1969, 91, 791; d) L. T.

Article Doi

New results in homoheptalene chemistry

DOI: 10.1002/1522-2675(20010516)84:5<1017::AID-HLCA1017>3.0.CO;2-L

Source and publish data:

Authors:

Article abstract of DOI:10.1002/1522-2675(20010516)84:5<1017::AID-HLCA1017>3.0.CO;2-L

Full text of DOI:10.1002/1522-2675(20010516)84:5<1017::AID-HLCA1017>3.0.CO;2-L

Products guided by the article

R&D Labs maybe for 78924-86-4

Relevant to this article

Hot Product