Benzoxepine formation by the 1,7 electrocyclisation of diene-conjugated carbonyl ylides: Studies on relative rates of cyclisation via intramolecular competition reactions

Article abstract of DOI:10.1039/a702962d

A series of reactions has been carried out using reactants of the type 11/12 in which carbonyl ylide cyclisation on to the substituent at the 6 position is in competition with cyclisation on to the unsubstituted phenyl group at the 2 position. The relative reactivity of the two groups, determined by measuring the product ratio 13:15 and 14:16, has been determined for a series of 6-substituted compounds. Alkenyl groups and the thiophene ring are found to be ca. 10-20 times more reactive than phenyl. In cases where the 6-substituent is a substituted aryl group it has been found that, unlike the analogous nitrile ylides, aromatic substituents have little effect on the reactivity of the ring. The selectivity is unaffected by the nature of the substituent on the terminal atom of the carbonyl ylide. Mechanistic studies using deuteriated reactants show that the cyclisation step is irreversible.

Full text of DOI:10.1039/a702962d

View Article Online / Journal Homepage / Table of Contents for this issue
Benzoxepine formation by the 1,7 electrocyclisation of diene-
conjugated carbonyl ylides: studies on relative rates of cyclisation via
intramolecular competition reactions
Donal F. O’Shea and John T. Sharp*
Department of Chemistry, University of Edinburgh, West Mains Road, Edinburgh, UK
EH9 3JJ
A series of reactions has been carried out using reactants of the type 11/12 in which carbonyl ylide
cyclisation on to the substituent at the 6 position is in competition with cyclisation on to the unsubstituted
phenyl group at the 2 position. The relative reactivity of the two groups, determined by measuring the
product ratio 13:15 and 14:16, has been determined for a series of 6-substituted compounds. Alkenyl
groups and the thiophene ring are found to be ca. 10–20 times more reactive than phenyl. In cases where
the 6-substituent is a substituted aryl group it has been found that, unlike the analogous nitrile ylides,
aromatic substituents have little effect on the reactivity of the ring. The selectivity is unaffected by the
nature of the substituent on the terminal atom of the carbonyl ylide. Mechanistic studies using
deuteriated reactants show that the cyclisation step is irreversible.
Table 1 Identification of 6-substituent in componds 9–12, 19, 22, and
8-substituent in 15, 16
Introduction
This work is concerned with the electrocyclisation reactions of
unsaturated 1,3-dipolar intermediates and in particular with
the 1,7 cyclisation of diene-conjugated systems 1, which pro-
vides a general route to unsaturated seven-membered heter-
ocycles 2.1 It describes further work on our investigations into
the structural factors which control the rates of these reactions.
The initial work2 was concerned with nitrile ylides and utilised
intramolecular competition reactions in the system 3 to deter-
mine the reactivity of a series of substituents at position 6
relative to that of the unsubstituted phenyl group at position 2.
The results, summarised in Table 8, showed that: (i) the reactiv-
ity was strongly dependent on the ‘double bond character’ of
the bond under attack, as would be expected; and (ii) the
reactivity of the phenyl groups was enhanced by both electron
withdrawing and electron donating aromatic substituents, par-
ticularly when at the meta position. The latter effect was
unexpected and contrasts with the behaviour of nitrile ylides
(Sustmann type I dipoles) in cycloaddition reactions, where the
reactivity of the dipolarophile is increased by electron with-
drawing groups but diminished by electron donating groups.
The next objective was to find out if the same pattern of reactiv-
ity was general to all 1,3-dipolar intermediates or, as for
cycloaddition reactions, it varied in a systematic way with the
nature of the 1,3-dipole.
This paper reports a study of carbonyl ylide cyclisation reac-
tions which almost exactly parallels the earlier work on nitrile
ylides. Carbonyl ylides were selected for several reasons: (i)
extensive earlier work on the synthetic and mechanistic aspects
of their 1,7 electrocyclisation reactions, largely from Eberbach’s
group, had shown that they cyclise effectively in systems where
both unsaturated elements of the conjugated diene system are
aromatic3 (e.g. 6 → 8) or heteroaromatic,4 and that the cyclis-
ation definitely proceeds via a formally conrotatory 1,7 electro-
cyclisation mechanism;3 and (ii) their use would extend the
reactivity studies to dipoles of the allyl type in contrast to the
nitrile ylides studied earlier which are of the propargyl–
allenyl type.
Compd.
Substituent
a
b
c
d
e
f
g
h
i
Phenyl
(E)-2-Phenylethenyl
2-Thienyl
3-Thienyl
3,5-Dichlorophenyl
3,5-Bis(trifluoromethyl)phenyl
3,5-Dimethylphenyl
3-Nitrophenyl
4-Fluorophenyl
4-Methylphenyl
4-Chlorophenyl
4-Methoxyphenyl
4-(Trifluoromethyl)phenyl
j
k
l
m
ylides 11, 12 were generated by the flash vacuum pyrolysis
(FVP) of the oxiranes 9, 10 at 625 ЊC. The background and
mechanistic assumptions implicit in this method were discussed
in the earlier work2 and, as there, the reactivity of a range of
unsaturated substituents (Table 1) incorporated at the 6 pos-
ition, relative to that of the phenyl group at the 2 position, has
been determined by measuring the ratio of the products 13:15
and 14:16. In each case the identity of one of the cyclisation
products, 15 or 16, was established by comparison with an
‘authentic’ sample.
Synthesis of the 3-[(2-phenyl-6-substituted)phenyl]oxiranes
9a–m, 10f,g as carbonyl ylide precursors and the ‘authentic’
dibenzoxepines 15a–m, 16f,g
Two series of oxiranes were studied, the major series 9a–m
(Scheme 1) had an ester as the R substituent while the minor
series 10f,g had a phenyl substituent. The unsaturated groups
incorporated at the 6 position, whose relative reactivity was to
be measured, are identified in Table 1. The preparative routes
are shown in Scheme 2. The key intermediate for the synthesis
of both series and for the ‘authentic’ dibenzoxepines 15a–m and
16f,g was 2-bromo-6-phenylbenzaldehyde 18. This was pre-
pared in good yield by the Bender and Hass method5 from the
known2 benzyl bromide 17. The ester-substituted compounds
9a–m were then prepared by two further steps, Suzuki coup-
ling with the appropriate aryl, heteroaryl or vinyl boronic acid
to give the substituted benzaldehydes 19a–m (Tables 2 and 3)
Results and discussion
The relative reactivity data was determined via the internal
competition reaction shown in Scheme 1, in which the carbonyl
J. Chem. Soc., Perkin Trans. 1, 1997
3025

View Article Online
(i)
(ii)
(iii)
9a–m
c
b
a
b
c
a
Ph
Br
Ph
Br
Ph
1
2
CH2Br
CHO
CHO
S
17
18
19
2
6
(iii) for 20
(iv) and (v) for 21
(iv)
S
–CH
+N
(v)
10f,g
CPh
Ph
Br
Ph
3
CH
0 °C, THF
S
O
CHPh
R
20 R = CO2Me
21 R = Ph
22
Ph
S
S
H
H
H
(vi)
N
N
H
Ph
Ph
4
5
Br
(ii)
15b–m; 16f,g
H
O
H
R
H
23 R = CO2Me
24 R = Ph
C
O
CHR
H
6
Scheme 2 Reagents and conditions (i) NaOEt–CH3CH(NO2)CH3; (ii)
ArB(OH)2 or vinylB(OH)2–Pd; (iii) ClCH2CO2Me–NaOMe; (iv)
(EtO)2P(O)CH2Ph–NaOEt; (v) MCPBA; (vi) FVP, 625 ЊC
FVP, 625 °C
Competitive cyclisation reactions of the carbonyl ylides 11a–m
and 12f,g
The overall course of the competition reactions is shown in
Scheme 1. The methods used are discussed in (1) below and in
more detail in the Experimental section. The results and the
mechanism of the reaction are discussed in (2) below.
H
R
[1,5]H
shift
R
H
O
H
O
H
H
8
7
(1) Experimental method. In a typical flash vacuum pyrolysis
experiment the oxirane 9 or 10 (ca. 0.5 mmol) was distilled
(150–180 ЊC) into an unpacked pyrolysis tube (625 ЊC) and the
products were collected in a cold trap at the exit. Preliminary
work had shown that the reactants could be distilled under high
vacuum without decomposition. The whole product mixture
was then dissolved in deuteriochloroform and analysed for the
13:15 or 14:16 ratio by 1H NMR spectroscopy and by HPLC.
The reaction mixtures were very clean and contained only the
expected products with, in some cases, a little residual oxirane.
They were then worked up and the dibenzoxepines were iso-
lated as a mixture by dry column flash chromatography. The
yields and product ratios are shown in Table 8. In the NMR
analyses the spectra were run at either 200 or 360 MHz at a
temperature of 328 K. The elevated temperature was required
because, in spectra obtained at 298 K, some of the peaks were
broadened due to ring inversion of the oxepine ring. The
spectra all showed a set of characteristic peaks (Table 9) for
each of the two isomers, with a singlet due to the proton on C-5
and a pair of doublets due to the methylene group at C-7. In
each mixture the peaks due to the 8-substituted isomer (15 or
16) were identified from the NMR spectrum of the ‘authentic’
sample prepared as discussed above, using peak enhancement
where necessary. The product ratio (Table 8) was determined
from the integrals of the 5-H absorptions and also measured
using HPLC. Control experiments showed that the isomer ratio
was not changed when the product mixture was passed through
the pyrolysis tube under the reaction conditions and also that
the isolated single isomers 15 were not isomerised into 13.
(2) The relative reaction rates and the mechanism of the cyclis-
ation. The results are shown in Table 8 alongside those for the
analogous nitrile ylide cyclisations.2 In comparing the two sets
of data it is important to be aware of the major differences
2
2
6
6
FVP, 625 °C
Ph
CH
+O
S
S
O
R
CHR
9 R = CO2Me
10 R = Ph
11 R = CO2Me
12 R = Ph
8
Ph
H
H
7
7
5
5
S
S
O
O
H
H
R
R
H
H
15 R = CO2Me
16 R = Ph
13 R = CO2Me
14 R = Ph
Scheme 1
and finally a Darzens reaction to give 9a–m (Tables 4 and 5).
The phenyl substituted analogues 10f,g were synthesised from
18 via Suzuki coupling to give 19, Wadsworth–Emmons olefin-
ation to give 22, and finally oxidation with m-chloroperbenzoic
acid. The ‘authentic’ dibenzoxepines 15b–m and 16f,g, required
for identification of the products of the competition reactions,
were prepared by FVP of the appropriate bromo-substituted
oxirane 20, 21 to give the 4-bromooxepines 23, 24 which were
then converted into the target compounds (Tables 6 and 7) by
Suzuki coupling.
3026
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
Table 2 Yields and physical data for the 2-phenyl-6-arylbenzaldehydes 19a–ma
Found (calc.) (%)
Reaction
time/h
Yield
(%)
Cryst.
Molecular
formula
Compd.
19a
19b
19c
solventb
Mp/ЊC
77–78
C
H
N
Mϩ
3
3
3
5
3
3
4
4
3
3
2
5
5
81
90
87
93
85
81
91
85
83
85
80
84
82
H
H
C19H14O
88.7
(88.4)
88.6
(88.7)
77.0
(77.2)
76.9
5.7
(5.4)
5.5
(5.6)
4.65
(4.55)
4.5
258.106 53
(258.10 477)
284.119 34
(284.120 12)
264.059 90
(264.060 89)
264.059 14
(264.060 89)
326.024 62
(326.026 52)
394.077 88
(394.079 23)
286.135 80
(286.135 77)
303.088 86
(303.089 54)
276.092 78
(276.095 04)
272.120 92
(272.120 12)
292.065 91
(292.065 49)
288.114 54
(288.115 03)
326.093 08
(326.091 85)
92–93
C21H16O
oil
C17H12OS
C17H12OS
C19H12Cl2O
C21H12F6O
C21H18O
19d
19e
H–E
H–E
H
76–77
(77.2)
69.6
(69.75)
64.25
(63.95)
88.5
(4.8)
3.8
123–124
131–132
109–110
124–125
112–113
87–88
(3.65)
2.95
(3.05)
6.25
(6.3)
4.5
19f
19g
19h
19i
H
(88.1)
75.3
H–E
H
C19H13NO3
C19H13FO
C20H16O
4.8
(4.6)
(75.2)
(4.3)
19j
H
88.45
(88.2)
77.8
(77.9)
83.55
(83.3)
73.4
6.0
(5.85)
4.5
(4.44)
5.85
(5.55)
4.1
19k
19l
H
97–98
C19H13ClO
C20H16O2
C20H13F3O
H–E
H–E
85–86
19m
87–88
(73.6)
(4.0)
a Compounds identified in Table 1. b H = hexane; E = diethyl ether.
in the reaction conditions; the nitrile ylide cyclisations were
carried out in a relatively polar solvent (THF) at 0 ЊC whereas
the carbonyl ylides were cyclised in the gas phase at 625 ЊC. The
results are similar in two respects but show one major differ-
ence. The first similarity is in the effect of the ‘double bond
character’ of the bond on to which cyclisation occurs [cases (b),
(c) and (d), Table 8]; in both cyclisations the order of reactivity
is as expected, i.e. olefin > thiophene > benzene. The differ-
ences, however, are much smaller for the carbonyl ylide cyclis-
ations, but this may simply reflect the lower selectivity expected
at the higher reaction temperature. The results for the para-
substituted aryl rings [cases (i)–(m)] are also similar for the two
cyclisations in that all the substituents have a slight activating
effect. The striking difference is in the results for the meta sub-
stituted aryl rings [cases (e)–(h)]. In the nitrile ylide cyclisations
all the meta substituents exerted a substantial activating effect
[8 to > 100 for cases (f)–(h)], whereas in the carbonyl ylide reac-
tions a small activation (1.3–1.5) was produced in two cases [(e)
3,5-dichloro- and (h) 3-nitro-] and a slight deactivation (0.7–
0.8) in the other two cases [(g) 3,5-dimethyl- and (f) 3,5-
bis(trifluoromethyl)-].
S
+
–
C
O
CHR
H
Ph
25
1,7
cycl.
S
S
H
H
R
CHR
C
H
O
H
O
Ph
Ph
H
27
26
[1,5]
H shift
S
2
S
S
S
S
4
3
3
2
4
S
S
R
H
H
H
O
Ph
Ph
H
Ph
H
N
N
28
Ph
Ph
H
H
Scheme 3
29
30
S
S
S
S
4
The difference in the effect of meta substituents for the two
dipoles is not easy to explain using the model suggested for
rationalising the nitrile ylide results.2 The course of the car-
bonyl ylide cyclisation is shown in Scheme 3; it is a two step
process in which the 1,7 cyclisation proceeds through the transi-
tion state 26 to the intermediate 27 which then undergoes a [1,5]
hydrogen shift to give the product. It is known3 that the 1,7
cyclisation step is formally conrotatory and it has been sug-
3
3
2
2
4
S
S
H
H
R
R
H
H
O
O
Ph
Ph
H
H
31
32
J. Chem. Soc., Perkin Trans. 1, 1997
3027

View Article Online
Table 3 NMR and IR spectroscopic data for the aldehydes 19a–ma
νmax/cmϪ1
(C᎐O)
᎐
Compd.
19a
δH
δC
7.35–7.65 (13 H, m, aromatic), 10.0 (1 H, s, CHO)
193.5 (CHO), 144.1 (C-1), 139.4 (C-2), 132.9 (C-1, phenyl), 1695
131.4, 130.2, 129.4, 128.0, 127.5 (aromatic CH signals)
194.1 (CHO), 146.7 (C-1), 138.5 (C-2), 138.3 (C-6), 137.0 1693
[C-1, (ethenyl phenyl)], 131,1 [C-1, (C-2 phenyl)], 132.3,
132.1, 129.8, 129.7, 128.4, 128.1 127.8, 127.7, 126.9, 126.8,
126.1 (remaining CH signals)
19b
7.12 (1 H, d, J 16.2, ethenyl CH), 7.25–7.79 (13 H, m,
aromatic), 8.1 (1 H, d, ethenyl CH), 10.03 (1 H, s, CHO)
19c
19d
19e
19f
7.1–7.7 (11 H, m, aromatic), 10.05 (1 H, s, CHO)
192.1 (C᎐O), 140.3 (C-1), 139.1 (C-6), 138.2 (C-2, 1692
᎐
thiophene), 134.9 (C-2), 133.2 [C-1, (C-2 phenyl)], 131.2,
130.0, 129.8, 129.5, 128.4, 128.1, 127.6, 127.5, 127.0
(aromatic CH signals)
7.14 (1 H, dd, J 5.3, 1.0, aromatic), 7.25–7.48 (9 H, m, 193.5 (CHO), 144.0 (C-1), 139.5 (C-6), 139.4 (C-2), 138.4 1693
aromatic), 7.54–7.69 (1 H, m, aromatic), 10.00 (1 H, s,
CHO)
[C-3, (C-6 thiophene)], 133.2 [C-1, (C-2 phenyl)], 131.4,
130.2, 130.0, 129.4, 129.2, 128.0, 127.5, 125.3, 124.2
(aromatic CH signals)
7.20–7.28 (2 H, m, aromatic), 7.32–7.50 (7 H, m,
aromatic), 7.57–7.65 (2 H, m, aromatic), 9.92 (1 H, s, [C-1, (C-6 Cl2C6H3)], 134.3 [C-3, (C-6 Cl2C6H3)], 132.4
CHO)
192.6 (CHO), 145.5 (C-1), 143.2 (C-6), 140.4 (C-2), 138.3 1691
[C-1, (C-2 phenyl)], 131.8, 130.9, 130.1, 129.6, 128.2, 128.0,
127.5, 127.3 (aromatic CH signals)
192.4 (CHO), 146.5 (C-1), 142.6 (C-6), 139.8 (C-2), 137.8 1690
{C-1, [C-6 3,5-(CF3)2C6H3]}, 131.9 [C-1, (C-2 phenyl)],
125.9 (CF3), 120.5 {C-3, [C-6 3,5-(CF3)2C6H3]}, 132.1,
131.2, 130.5, 129.8, 129.3, 128.4, 128.2, 121.0 (remaining
CH aromatic signals)
7.3–7.9 (11 H, m, aromatic), 9.9 (1 H, s, CHO)
19g
19h
2.4 (6 H, s, CH3), 7.0–7.5 (11 H, m, aromatic), 10.0 (1 H, 193.6 (CHO), 144.7 (C-1), 143.6 (C-6), 139.7 (C-2), 139.2 1692
s, CHO)
{C-3 [3,5-(CH3)2C6H3]}, 137.5 {C-1 [3,5-(CH3)2C6H3]},
132.9 [C-1, (C-2 phenyl], 21.1 (CH3), 131.2, 130.0, 129.3,
129.2, 128.7, 127.3 [remaining CH aromatic signals,
(overlap of 2 CH signals)]
7.32–7.73 (10 H, m, aromatic), 8.20–8.27 (2 H, m, 192.6 (CHO), 147.6 [C-3, (C-6 3-NO2C6H4)], 146.1 (C-1), 1695
aromatic), 9.92 (1 H, s, CHO)
142.1 (C-6), 140.4 (C-2), 138.0 [C-1, (C-6 3-NO2C6H4)],
132.2 [C-1, (C-2 phenyl)], 135.2, 132.0, 131.0, 130.4, 130.1,
129.7, 128.6, 128.3, 128.1, 123.9, 122.7 (CH aromatic
signals)
19i
19j
7.07–7.14 (2 H, m, aromatic), 7.24–7.46 (9 H, m,
aromatic), 7.55–7.61 (1 H, m, aromatic), 9.92 (1 H, s,
CHO)
2.47 (3 H, s, CH3), 7.3–7.65 (12 H, m, aromatic), 10.04 (1
H, s, CHO)
1698
193.4 (CHO), 144.1 (C-1), 143.9 (C-6), 139.5 (C-2), 137.2 1694
[C-4 (4-CH3C6H4)], 136.3 [C-1 (4-CH3C6H4)], 132.9 [C-1,
(C-2 phenyl)], 21.0 (CH3), 131.2, 130.1, 129.9, 129.3, 128.7,
127.9, 127.3 [remaining CH aromatic signals, (overlap of
one CH signal)]
19k
19l
7.28–7.64 (12 H, m, aromatic), 9.95 (1 H, s, CHO)
193.1 (CHO), 144.8 (C-1), 142.4 (C-6), 138.9 (C-2), 138.3 1695
[C-4 (4-ClC6H4)], 133.5 [C-1 (4-ClC6H4)], 132.8 [C-1, (C-2
phenyl)], 131.5, 130.6, 130.4, 130.2, 129.5, 128.1, 127.7
(aromatic CH signals, overlap of one CH signal)
3.86 (3 H, s, OCH3), 6.95–7.02 (2 H, m, aromatic), 7.29–
7.62 (10 H, m, aromatic), 9.98 (1 H, s, CHO)
193.5 (CHO), 159.1 [C-4 (4-CH3OC6H4)], 143.9 (C-1),
143.8 (C-6), 139.6 (C-2), 132.9 [C-1, (C-2 phenyl)], 131.4
[C-1 (4-CH3OC6H4)], 113.5 [C-3 (4-CH3OC6H4)], 55.1
(OCH3), 131.3, 130.6, 130.2, 129.8, 129.3, 127.9, 127.3
(remaining aromatic CH signals)
1696
19m
7.32–7.49 (8 H, m, aromatic), 7.59–7.70 (4 H, m,
aromatic), 9.94 (1 H, s, CHO)
192.9 (CHO), 145.2 (C-1), 143.9 (C-6), 141.9 (C-2), 138.5 1692
[C-4 (C-6 4-FC6H4)], 132.6 [C-1, (C-2 phenyl)], 129.1 [C-1,
(C-6 4-FC6H4)], 124.8 (CF3), 131.7, 130.7, 130.3, 129.6,
129.5, 128.2, 127.9 (aromatic CH signals)
a Compounds identified in Table 1
gested that it, like the analogous nitrile ylide reaction, goes
through a helical transition state. However it is not known
whether the transition state is early or late. In the analogous
nitrile ylide cyclisations2 it was suggested that the transition
state was late and that the intermediate corresponding to 27
could therefore be taken as a model of the transition state and
that the latter would therefore be subject to similar stabilising/
destabilising effects by the substituents. It was suggested that
the canonical forms 29, 30 would make a significant contribu-
tion to the resonance stabilisation of the intermediate since
both contain an intact benzene ring and that these structures
would be further stabilised by aromatic substituents in the
upper ring with, respectively, electron withdrawing or donating
properties. The substituents would similarly serve to stabilise
the transition state for the cyclisation, assuming it is ‘product’-
like, and so enhance the reactivity of the ring. Such substituents
would have a maximum effect when conjugated with the charge
to be delocalised, i.e. in the 2 or 4 positions (as in the cyclisation
of the meta-substituted cases) and a lesser effect when in the 3
position (as in the cyclisation of the para-substituted cases).
The closest analogues to 29, 30 for the intermediate 27 in the
carbonyl ylide cyclisation are shown in structures 31, 32. Of
these, 31 would be expected to be the more stable because
of electron donation by the oxygen atom adjacent to the carbon
bearing the positive charge. By the same argument used in the
nitrile ylide case it would be predicted that electron withdraw-
ing substituents at the 2 and/or 4 positions would stabilise 31
and electron donating groups at these positions would stabilise
32, and in both cases thus enhance the reactivity of the ring.
Experimentally this is found not to be true. The failure of this
3028
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
Table 4 Yields and physical data for the 2-methoxycarbonyl-3-[(2-phenyl-6-substituted)phenyl]oxiranes 9a–ma
Found (calc.) (%)
Reaction
time/h
Yield
(%)
Cryst.
Molecular
formula
Compd.
9a
solventb
Mp/ЊC
C
H
N
Mϩ
5
3
85
81
78
76
82
89
84
74
81
76
78
77
83
H–E
H
153–154
103–104
131–132
141–142
121–122
112–113
68–70
C22H18O3
330.1257
(330.1256)
356.1422
(356.1412)
336.0812
(336.0820)
336.0823
(336.0820)
398.0509
(398.0477)
466.1012
(466.1004)
358.1539
(358.1569)
375.1095
(375.1107)
348.1156
(348.1161)
344.1410
(344.1412)
364.0870
(364.0866)
360.1377
(360.1362)
398.1095
(398.1130)
9b
C24H20O3
80.6
(80.9)
71.3
(71.45)
71.7
(71.45)
66.2
(66.3)
62.2
5.4
(5.6)
4.9
(4.75)
5.1
(4.75)
4.0
(4.0)
3.7
9c
3
H–E
H–E
H–E
H–E
H–E
H–E
H–E
H–E
H–E
H–E
H–E
C20H16O3S
C20H16O3S
C22H16Cl2O3
C24H16F6O3
C24H22O3
9d
12
5
9e
9f
4
(61.8)
(3.4)
9g
6
9h
4
124–125
137–138
127–128
156–157
180–181
176–177
C22H17NO5
C22H17FO3
C23H20O3
70.25
(70.4)
75.45
(75.85)
80.1
(80.2)
72.3
(72.4)
76.3
(76.65)
69.1
4.5
(4.5)
4.6
4.0
(3.75)
9i
2
(4.85)
5.85
(5.8)
4.9
(4.65)
5.55
(5.55)
4.35
(4.25)
9j
24
2
9k
9l
C22H17ClO3
C23H20O4
10
6
9m
C23H17F3O3
(69.35)
a Compounds identified in Table 1. b H = hexane; E = diethyl ether.
prediction for carbonyl ylides may be because the rationalis-
ation of substituent effects is wholly wrong in the nitrile ylide
case, or it may be due to the difference in the nature of the 1,3-
dipoles or in the reaction conditions. Several possible explan-
ations in the latter category are discussed below.
35/37 = (xkϪ1 ϩ kH)/(kϪ1 ϩ kH)
(1)
ratio, where x = kH/kD. Thus the ratio will be 1 if kϪ1 = 0 and
will tend to x as kϪ1 ӷ kH.
Cyclisation of the carbonyl ylide 33 was carried out under
the usual conditions and the mixture of isomers 35 and 37 was
isolated by flash column chromatography. Mass spectroscopy
of this mixture showed that no deuterium had been lost in
either the reaction or work-up. The 2H NMR spectrum showed
that the deuterium in the product mixture was located only on
C-7 (δD 4.37) (compound 37) and in the aromatic ring (δD 7.58)
(compound 35). The integral ratio of these peaks and hence the
product ratio was 0.99 0.02 (average from three cyclisations).
Thus the cyclisation step is also irreversible in this reaction and
the difference in the substituent effects must be due to some
other cause.
Since neither of the above hypotheses provide an explan-
ation, then it may be that the difference in substituent effects
is simply due to the difference in reaction conditions, i.e.
the absence of solvation in the carbonyl ylide reaction. This
could in principle be investigated experimentally by generat-
ing the carbonyl ylides in solution via an alternative route.
The choice of such a route, however, is not straight-
forward but it is hoped to do this in future work. More
interestingly, it is possible that the difference is due to the
intrinsic difference in unsaturation between the two types of
1,3-dipole. This affects the extent of charge delocalisation in
the intermediate, i.e. 31/32 for the allyl type dipole (cf. 29, 30
for the propargyl–allenyl dipole). Again, assuming a late
transition state, the absence of charge delocalisation on to the
‘attacking’ terminal carbon of the 1,3-dipole of the carbonyl
ylide would diminish the effect of any coulombic assistance2
to cyclisation produced by the polarising effect of the
substituents.
Effect of the nature of the substituent on the terminal carbon
of the carbonyl ylide.—One plausible possibility was that the
presence of the electron withdrawing ester group on the attack-
ing carbon of the carbonyl ylide was affecting or negating the
predicted substituent effect. This was tested by cyclising the two
compounds 12f and 12g, which had a phenyl substituent in
place of the ester in the main series. These examples had
respectively 3,5-bis(trifluoromethyl) and 3,5-dimethyl substitu-
ents in the aromatic ring, both of which had a moderate activat-
ing effect on the nitrile ylide cyclisation and a mild deactivating
effect on the cyclisation of 11f,g. The relative reactivities (Table
8) were found to be unaffected by the change of substituent and
this explanation was therefore discarded.
Reaction mechanism.—Another, more fundamental, possibil-
ity was that the mechanisms of the two reactions may differ in
the identity of the rate (and hence product ratio) determining
step. As discussed above these cyclisation reactions, e.g.
25 → 28, involve two steps; 1,7 electrocyclisation, followed by
a [1,5] sigmatropic hydrogen migration which leads to the prod-
uct. Mechanistic study of the analogous nitrile ylide cyclisation
has shown that the cyclisation step is irreversible1c and hence
that the selectivity is exerted in this step. In this work on car-
bonyl ylide cyclisation it has been shown by control reactions
that the overall process is irreversible, i.e. the isomers 13 and 15
do not interconvert under the cyclisation conditions. However
this does not rule out the possibility that the first step might be
reversible in this case due to the high temperature of the reac-
tion, and thus that the product ratio would be affected by the
rate constants of both steps. This has been investigated by one
of the methods developed in earlier work using the deuteriated
reactant 33 (Scheme 4).6 This was generated in the usual way
from the appropriate deuteriated oxirane (98% monodeuteri-
ated). As discussed earlier5 the application of the steady state
approximation to 34 and 36 leads to Eqn. (1) for the product
The main practical conclusion which can be drawn is that the
rate of the 1,7 carbonyl ylide cyclisations is little affected by
either the olefinic/aromatic character of the γ,δ bond or
whether it is electron rich or electron poor. This fact, combined
with the use of the Suzuki coupling to provide an easy route to
J. Chem. Soc., Perkin Trans. 1, 1997
3029

View Article Online
Table 5 NMR and IR spectroscopic data for the 2-methoxycarbonyl-3-[(2-phenyl-6-substituted)phenyl]oxiranes 9a–ma
νmax/cmϪ1
(C᎐O)
᎐
Compd.
δH
δC
9a
2.72 (1 H, d, J 2.0, 2-H), 3.49 (3 H, s, OCH3), 4.43 (1 H,
d, J 2.0, 3-H), 7.32–7.51 (13 H, m, aromatic)
168.5 (CO), 143.1 (q), 140.4 (q), 130.6 (q), 57.7 (C-3), 55.7
(C-2), 51.9 (OCH3), 129.0, 128.8, 128.3, 128.2, 127.2
(aromatic CH signals)
1695
9b
9c
9d
9e
9f
2.98 (1 H, d, J 2.0, 2-H), 3.60 (3 H, s, OCH3), 4.38 (1 H,
d, J 2.0, 3-H), 7.08 (1 H, d, J 16.1, ethenyl CH), 7.21–7.66
(13 H, m, aromatic), 7.70 (1 H, d, J 16.1, ethenyl CH)
168.8 (CO), 142.6 (q), 140.3 (q), 138.2 (q), 137.0 (q), 130.0
(q), 57.1 (C-3), 55.2 (C-2), 52.1 (OCH3), 131.4, 128.8,
128.6, 128.5, 128.3, 127.8, 127.1, 126.7, 125.6, 124.7
(remaining CH signals), (overlap of one CH signal)
168.6 (CO), 143.3 (q), 141.2 (q), 140.3 (q), 135.5 (q), 131.0
(q), 57.4 (C-3), 56.0 (C-2), 52.0 (OCH3), 129.6, 128.7,
128.4, 128.3, 127.3, 127.0, 125.9 (aromatic CH signals,
overlap of 2 CH signals)
1740
2.8 (1 H, d, J 2.0, 2-H), 3.55 (3 H, s, OCH3), 4.48 (1 H, d,
J 2.0, 3-H), 7.09–7.24 (3 H, m, aromatic), 7.30–7.49 (8 H,
m, aromatic)
1739
1740
1740
1745
1743
1742
1739
1736
1738
1743
1740
2.78 (1 H, d, J 2.0, 2-H), 3.55 (3 H, s, OCH3), 4.43 (1 H,
d, J 2.0, 3-H), 7.23–7.25 (1 H, m, aromatic), 7.30–7.46 (q), 57.4 (C-3), 55.6 (C-2), 52.0 (OCH3), 129.1, 128.8,
(10 H, m, aromatic)
168.6 (C᎐O), 143.1 (q), 140.5 (q), 140.4 (q), 137.7 (q), 130.5
᎐
128.7, 128.4, 128.3, 128.2, 127.1, 125.4, 123.1 (aromatic
CH signals)
2.74 (1 H, d, J 1.9, 2-H), 3.54 (3 H, s, OCH3), 4.34 (1 H,
d, J 1.9, 3-H), 7.29–7.50 (11 H, m, aromatic)
168.2 (C᎐O), 143.5 (q), 143.4 (q), 140.3 (q), 140.0 (q), 134.7
᎐
(q), 130.5 (q), 57.1 (C-3), 55.5 (C-2), 52.1 (OCH3), 129.8,
128.8, 128.7, 128.5, 128.3, 127.4, 127.3 (aromatic CH
signals), (overlap of one CH signal)
2.72 (1 H, d, J 2.0, 2-H), 3.5 (3 H, s, OCH3), 4.32 (1 H, d,
J 2.0, 3 H), 7.29–7.55 (8 H, m, aromatic), 7.88–7.91 (3 H,
m, aromatic)
168.0 (C᎐O), 147.2 (q), 143.6 (q), 142.7 (q), 139.9 (q), 139.8
᎐
(q), 130.7 (q), 121.1 (CF3), 56.9 (C-3), 55.5 (C-2), 52.0
(OCH3), 130.2, 129.1, 128.9, 128.7, 128.4, 127.6, 125.3
(aromatic CH signals), (overlap of one CH signal)
9g
9h
9i
2.46 [6 H, s, (CH3)2C6H3], 2.82 (1 H, d, J 2.0, 2-H), 3.57 (3 168.4 (C᎐O), 143.1 (q), 142.8 (q), 140.4 (q), 140.2 (q), 137.4
H, s, OCH3), 4.55 (1 H, d, J 2.0, 3-H), 7.13 (3 H, br s,
aromatic), 7.36–7.54 (8 H, m, aromatic)
᎐
(q), 130.3 (q), 57.5 (C-3), 55.5 (C-2), 51.6 (OCH3), 21.0
(CH3), 129.1, 128.7, 128.6, 128.1, 127.7, 127.2, 126.9, 126.4
(aromatic CH signals)
2.72 (1 H, d, J 2.0, 2-H), 3.47 (3 H, s, OCH3), 4.37 (1 H,
d, J 2.0, 3-H), 7.30–7.80 (10 H, m, aromatic), 8.21–8.30 (2
H, m, aromatic)
168.1 (C᎐O), 147.9 (q), 143.4 (q), 142.2 (q), 140.4 (q), 139.8
᎐
(q), 132.9 (q), 57.1 (C-3), 55.6 (C-2), 52.0 (OCH3), 134.8,
130.6, 129.7, 129.1, 128.7, 128.6, 128.3, 127.5, 123.6, 122.1
(aromatic CH signals)
2.71 (1 H, d, J 2.0, 2-H), 3.50 (3 H, s, OCH3), 4.36 (1 H,
d, J 2.0, 3-H), 7.09–7.20 (4 H, m, aromatic), 7.24–7.50 (8
H, m, aromatic)
168.5 (C᎐O), 164.1 (q), 143.1 (q), 142.0 (q), 140.3 (q), 136.5
᎐
(q), 130.5 (q), 115.0 (q), 57.5 (C-3), 55.7 (C-2), 51.9
(OCH3), 130.4, 130.3, 129.1, 128.7, 128.4, 128.2, 127.3
(remaining aromatic CH signals)
9j
2.42 (3 H, s, 4-CH3C6H4), 2.71 (1 H, d, J 2.0, 2-H), 3.48 (3 168.5 (C᎐O), 143.1 (q), 143.0 (q), 140.5 (q), 137.5 (q), 136.8
H, s, OCH3), 4.42 (1 H, d, J 2.0, 3 H), 7.24–7.46 (12 H, m,
aromatic)
᎐
(q), 130.5 (q), 57.7 (C-3), 55.6 (C-2), 51.7 (OCH3), 20.9
(CH3), 129.0, 128.8, 128.7, 128.6, 128.2, 128.1, 127.1
(aromatic CH signals), (overlap of one signal)
9k
9l
2.71 (1 H, d, J 2.0, 2-H), 3.51 (3 H, s, OCH3), 4.38 (1 H,
d, J 2.0, 3-H), 7.28–7.50 (12 H, m, aromatic)
168.4 (C᎐O), 143.1 (q), 141.8 (q), 140.2 (q), 138.9 (q), 133.3
᎐
(q), 130.5 (q), 57.5 (C-3), 55.6 (C-2), 52.0 (OCH3), 130.1,
129.2, 128.9, 128.7, 128.4,128.3, 128.2, 127.3 (aromatic CH
signals)
2.7 (1 H, d, J 2.0, 2-H), 3.50 (3 H, s, OCH3), 3.86 (3 H, s, 168.6 (C᎐O), 158.8 (q), 143.0 (q), 142.8 (q), 140.5 (q), 132.8
4-CH3OC6H4), 4.42 (1 H, d, J 2.0, 3-H), 6.98 (2 H, d, J
8.8, aromatic), 7.30–7.48 (10 H, m, aromatic)
᎐
(q), 130.6 (q), 57.7 (C-3), 55.7 (C-2), 55.0 (4-CH3OC6H4),
51.8 (OCH3), 129.8, 129.0, 128.7, 128.6, 128.3, 128.2,
127.1, 113.6 (aromatic CH signals)
9m
2.71 (1 H, d, J 2.0, 2-H), 3.48 (3 H, s, OCH3), 4.38 (1 H,
d, J 2.0, 3-H), 7.30–7.57 (10 H, m, aromatic), 7.70–7.73 (2
H, m, aromatic)
168.4 (C᎐O), 144.2 (q), 143.3 (q), 141.5 (q), 140.1 (q), 136.6
᎐
(q), 125.2 (CF3), 57.4 (C-3), 55.5 (C-2), 52.0 (OCH3), 129.6,
129.1, 128.9, 128.7, 128.5, 128.3, 127.4 (aromatic CH
signals)
a Compounds identified in Table 1.
the reactants, makes this a wide ranging and effective synthetic
route to fused oxepines.
sodium diphenyl ketyl immediately before use. 1,2-Dimethoxy-
ethane (DMF) was passed through a column of activated
alumina immediately before use. The arylboronic acids used in
the coupling reactions were all obtained from either Aldrich
Chemical Company or Lancaster Synthesis.
Experimental
NMR spectra were run as solutions in deuteriochloroform at
298 K unless otherwise stated. Chemical shifts are recorded as δ
values and J values are given in Hz. In the 13C spectra, carbon
multiplicity was established by DEPT. Mass spectra were
obtained using electron ionisation at 70 eV. Preparative chro-
matography was carried out by the ‘dry column flash’ tech-
nique7 using silica gel (15 µm, Fluka Kieselgel GF254) and
eluting solvents based on hexane admixed with diethyl ether.
Ether refers to diethyl ether. Evaporation of solvents indicates
evaporation under reduced pressure using a rotary evaporator.
All drying of solutions was done with anhydrous magnesium
sulfate.
General method for flash vacuum pyrolysis (FVP)
The appropriate oxirane was distilled (150–180 ЊC) at 1 × 10Ϫ4
mmHg through a furnace tube (35 × 2.5 cm) which was main-
tained at 625 ЊC. The product was collected in a U-tube which
was cooled by liquid nitrogen and situated at the exit of the
furnace.
Preparation of the 2-phenyl-6-substituted-benzaldehydes 19a–m
2-Bromo-6-phenylbenzaldehyde 18. Sodium (0.74 g, 0.032
mol) was dissolved in absolute ethanol (100 cm3), 2-nitro-
propane (3.36 g, 0.037 mol) was added and the solution was
stirred for 1 h. 2-Bromo-6-phenylbenzylbromide2 17 (3.82 g,
0.012 mol) was added and stirring continued for 72 h. The
Solvents, reagents and starting materials
Tetrahydrofuran (THF) was distilled under nitrogen from
3030
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
Table 6 Yields and physical data for the 5-methoxycarbonyl-8-substituted-5,7-dihydrodibenzo[c,e]oxepines 15b–ma
Found (calc.) (%)
Reaction
time/h
Yield
(%)
Cryst.
Molecular
formula
Compd.
15b
15c
solventb
Mp/ЊC
135–136
117–118
oil
C
H
Mϩ
24
24
24
8
86
80
76
84
65
74
77
67
80
62
80
91
H
C24H20O3
81.1
(80.9)
71.1
5.9
(5.6)
4.5
356.1436
(356.1412)
336.0806
(336.0820)
336.0806
(336.0820)
398.0480
(398.0477)
446.1027
(446.1004)
358.1568
(358.1569)
375.1084
(375.1107)
348.1160
(348.1162)
344.1410
(344.1413)
364.0888
(364.0866)
360.1355
(360.1362)
398.1130
(398.1130)
H–E
C20H16O3S
C20H16O3S
C22H16Cl2O3
C24H16F6O3
C24H22O3
(71.4)
(4.75)
15d
15e
H–E
87–90
66.6
(66.2)
62.1
4.05
(4.0)
3.3
15f
4
oil
(61.8)
(3.4)
15g
15h
15i
8
oil
8
H–E
H
115–116
119–120
169–170
155–156
138–139
139–140
C22H17NO5
C22H17FO3
C23H20O3
4
15j
2
H
79.9
6.0
(5.8)
4.9
(4.65)
5.6
(80.2)
72.55
(72.5)
76.3
15k
15l
3
H
C22H17ClO3
C23H20O4
2
H–E
H–EA
(76.65)
(5.55)
15m
4
C23H17F3O3
a Compounds identified in Table 1. b H = hexane; E = diethyl ether; EA = ethyl acetate.
ethanol was evaporated and the residue was dissolved in ether
(50 cm3). The solution was washed with water (2 × 50 cm3),
dried and the solvent was evaporated. Dry flash chrom-
atography (silica, hexane–ether, 80:20) of the residue gave 6-
bromo-2-phenylbenzaldehyde as a yellow oil (2.3 g, 82%) which
solidified on standing, mp 51–52 ЊC (Found: C, 60.2; H, 3.5.
C13H9BrO requires C, 60.0; H, 3.45%) (HRMS: found Mϩ,
259.9814. C13H979BrO requires M, 259.9837); δH 7.23–7.48 (6 H,
m, aromatic), 7.54–7.70 (2 H, m, aromatic) and 9.97 (1 H, s,
CHO); δC 191.8 (CHO), 146.4 (q, C-1), 137.7 (q, C-2), 122.5 (q,
C-6), 128.6 (q, C-1, phenyl), 133.2, 132.5, 130.0, 129.3, 128.2
and 128.0 (remaining CH aromatics); νmax(Nujol)/cmϪ1 1706
2,6-Diphenylbenzaldehyde 19a. 2-Bromo-6-phenylbenzalde-
hyde 18 (1.5 g, 5.75 mmol) and tetrakis(triphenylphosphine)-
palladium (0.2 g, 3%) in 1,2-dimethoxyethane (DME) (15 cm3)
were stirred for 20 min. A solution of phenylboronic acid (0.73
g, 6.0 mmol) and sodium carbonate (0.63 g, 6 mmol) in water
(10 cm3) was added and the mixture heated at reflux for 3 h.
After evaporation of the DME, dichloromethane (50 cm3) was
added and the organic layer was separated, dried and filtered
through a pad of activated alumina. The solvent was evapor-
ated and the residue was crystallised from hexane to give 2,6-
diphenylbenzaldehyde (1.2 g, 81%).
(C᎐O).
Preparation of the 2-(methoxycarbonyl)-3-[(2-phenyl-6-
substituted)phenyl]oxiranes 9a–m
᎐
Compounds 19a–m were prepared from the 2-bromo-6-
phenylbenzaldehyde 18 using the general method given in detail
for 2,6-diphenylbenzaldehyde 19a. Reaction times, yields and
the physical properties of the products are given in Table 2 and
their spectroscopic properties in Table 3.
Compounds 9a–m were prepared from the corresponding 2-
phenyl-6-substituted-benzaldehydes 19a–m using the general
method given in detail for 2-(methoxycarbonyl)-3-[(2,6-
diphenyl)phenyl]oxirane 9a. Reaction times, yields and the
physical properties of the products are given in Table 4 and
their spectroscopic properties in Table 5.
kH
D
D
2-(Methoxycarbonyl)-3-[(2,6-diphenyl)phenyl]oxirane 9a. A
solution of 2,6-diphenylbenzaldehyde 19a (0.75 g, 2.90 mmol)
and methyl chloroacetate (0.94 g, 8.66 mmol) in dry THF (25
cm3), under dry nitrogen, was treated with sodium methoxide
(0.40 g, 7.40 mmol) and heated under reflux for 5 h. The THF
was evaporated and the residue was dissolved in dichloro-
methane (50 cm3) and washed with water (2 × 20 cm3). The
dichloromethane solution was dried and passed through a pad
of activated alumina. Evaporation of the solvent and crystallis-
ation of the residue from hexane–ether gave the product 9a as a
white crystalline solid (0.82 g, 85%).
H
CO2Me
CO2Me
H
H
O
O
H
H
H
H
H
k1
k–1
34
35
D
C
O
CHCO2Me
k1
k–1
33
Preparation of the 5-methoxycarbonyl-8-substituted-5,7-
dihydrodibenz[c,e]oxepines 15b–m
Compounds 15b–m were prepared by the route shown in
Scheme 2. The general method used for the final step is given in
detail for the first example 15b. Reaction times, yields and the
physical properties of the products 15b–m are given in Table 6,
and their spectroscopic properties in Table 7.
kD
H
H
D
CO2Me
CO2Me
H
H
O
O
H
H
D
36
37
2-Methoxycarbonyl-3-[(2-bromo-6-phenyl)phenyl]oxirane 20.
This compound was prepared from 2-bromo-6-phenyl-
Scheme 4
J. Chem. Soc., Perkin Trans. 1, 1997
3031

View Article Online
Table 7 NMR and IR spectroscopic data for the 5-methoxycarbonyl-8-substituted-5,7-dihydrobenzoxepines 15b–ma
νmax/cmϪ1
(C᎐O)
᎐
b
c
Compd.
δH
δC
15b
3.50 (3 H, s, OCH3), 4.62 (2 H, br s, 7-Hax,eq), 5.09 (1 H, s,
5-H), 7.11 (1 H, d, J 16.1, ethenyl CH), 7.22–7.68 (13 H,
m, aromatic);
172.0 (C᎐O), 141.3 (q), 141.0 (q, broad), 137.5 (q), 137.1
1732
᎐
(q), 136.9 (q), 132.9 (CH), 132.6 (CH), 132.0 (q, broad),
129.4, 129.0, 128.9, 128.5, 128.4, 127.9, 127.8, 127.3, 126.6,
126.2 (CH), 62.6 (C-7), 51.9 (OCH3);d
([2H6]DMSO) 3.20 (3 H, s, CH3), 4.28 (1 H, d, J 12.2, 7-
([2H6]DMSO) 171.5 (C᎐O), 141.3 (q), 140.0 (q), 137.1 (q),
᎐
Hax), 4.75 (1 H, d, 7-Heq), 5.22 (1 H, s, 5-H), 6.77–7.69 (14 133.7 (q), 132.7 (CH), 132.3 (CH), 132.1 (q), 129.6, 129.4,
H, m, aromatic/phenyl ethenyl CH)
129.0, 128.8, 128.0, 127.0, 126.9, 126.4, 126.2, 125.5 (CH),
76.2 (C-5), 61.7 (C-7), 51.5 (OCH3)
15c
15d
15e
15f
3.43 (3 H, s, OCH3), 4.40 (1 H, d, J 11.5, 7-Hax), 4.51 (1 172.0 (C᎐O, br), 141.7 (q), 141.1 (q), 135.3 (q), 133.7 (q, br), 1729
᎐
H, d, 7-Heq), 5.17 (1 H, s, 5-H), 7.12–7.16 (1 H, m,
aromatic), 7.36–7.61 (9 H, m, aromatic)
132.5 (q), 131.8 (q, Br), 130.3, 129.5, 128.7, 128.6, 128.0,
127.8, 127.6, 127.0, 125.7 (overlap of one aromatic CH
signal), 64.1 (C-7), 51.9 (OCH3)d
3.41 (3 H, s, OCH3), 4.35 (1 H, d, J 12.0, 7-Hax), 4.56 (1 172.1 (C᎐O, br), 141.6 (q), 140.6 (q), 137.4 (q), 137.0 (q), 1731
᎐
H, d, 7-Heq), 5.17 (1 H, s, 5-H), 7.29–7.79 (10 H, m, 133.7 (q), 133.4 (q), 129.4, 128.9, 128.7, 128.4, 128.0,
aromatic)
126.5, 126.1, 125.9, 123.6, 119.6 (CH), 64.2 (C-7), 51.8
(OCH3)d
3.39 (3 H, s, OCH3), 4.23 (1 H, d, J 11.9, 7-Hax), 4.45 (1 171.9 (C᎐O, br), 143.1 (q), 141.5 (q), 139.8 (q), 134.8 (q), 1733
᎐
H, d, Heq), 5.18 (1 H, s, 5-H), 7.31–7.60 (10 H, m, 134.7 (q), 132.6 (q, br), (one quaternary signal not seen),
aromatic)
129.6, 129.5, 128.9, 128.7, 128.2, 127.8, 127.7, 127.5, 127.4
(CH), 77.1 (C-5), 63.6 (C-7), 51.9 (OCH3)
3.36 (3 H, s, OCH3), 4.20 (1 H, d, J 12.0, 7-Hax), 4.51 (1 171.9 (C᎐O, br), 142.3 (q), 141.8 (q), 139.4 (q), 139.2 (q), 1736
᎐
H, d, 7-Heq), 5.24 (1 H, s, 5-H), 7.38–7.62 (8 H, m,
aromatic), 7.94 (1 H, br s, aromatic), 8.28 (1 H, br s,
aromatic)
133.7 (q, br), 132.5 (q), 131.8 (q), 129.7, 129.6, 129.1, 128.8,
128.2, 127.9, 127.0 (overlap of two aromatic CH peaks)
121.1 (CF3), 63.6 (C-7), 51.8 (OCH3)d
([2H6]DMSO) 172.2 (C᎐O), 142.6 (q), 141.6 (q), 139.3 (q),
᎐
138.9 (q), 134.1 (q), 132.4 (q), 131.3 (q), 130.3, 130.1, 130.0,
129.9, 121.2 (CF3), 129.8, 129.5, 129.0, 128.3, 128.1 (CH),
76.8 (C-5), 63.4 (C-7), 51.6 (OCH3)
15g
2.42 (6 H, s, CH3), 3.48 (3 H, s, OCH3), 4.35 (1 H, d, J 171.9 (C᎐O, br), 142.8 (q), 141.1 (q), 140.1 (q), 137.9 (q), 1730
᎐
11.6, 7-Hax), 4.50 (1 H, d, 7-Heq), 5.19 (1 H, s, 5-H), 7.36– 137.6 (q), 133.4 (q, br), 132.0 (q, br), 129.9, 129.3, 128.8,
7.61 (10 H, m, aromatic)
128.4, 128.3, 128.1, 127.2, 126.5, 124.9, 63.9 (C-7), 51.8
(OCH3), 21.3 (CH3)d
15h
15i
15j
3.49 (3 H, s, CH3), 4.23 (1 H, d, J 11, 7-Hax), 4.41 (1 H, d,
7-Heq), 5.21 (1 H, s, 5-H), 7.25–7.71 (9 H, m, aromatic),
8.17–8.27 (2 H, m, aromatic)
3.42 (3 H, s, CH3), 4.26 (1 H, d, J 11.8, 7-Hax), 4.45 (1 H,
d, 7-Heq), 5.18 (1 H, s, 5-H), 7.20–7.69 (11 H, m,
aromatic)
1733
1738
1735
2.43 (3 H, s, CH3), 3.45 (3 H, s, OCH3), 4.33 (1 H, d, J 172.0 (C᎐O, br), 142.5 (q), 141.3 (q), 140.5 (q, br), 137.3
11.7, 7-Hax), 4.49 (1 H, d, 7-Heq), 5.18 (1 H, s, 5-H), 7.30– (q), 136.9 (q), 133.4 (q, br), 132.0 (q, br), 129.9, 129.4,
᎐
7.60 (11 H, m, aromatic)
129.3, 128.9, 128.7, 128.6, 128.4, 128.1, 126.5, 77.1 (C-5,
br), 64.1 (C-7), 51.8 (OCH3), 21.1 (CH3)
15k
15l
3.41 (3 H, s, OCH3), 4.25 (1 H, d, J 11.7, 7-Hax), 4.44 (1 172.0 (C᎐O, br), 141.4 (q), 141.3 (q), 141.0 (q, br), 138.6
H, d, 7-Heq), 5.18 (1 H, s, 5-H), 7.35–7.65 (11 H, m, (q), 133.4 (q), 132.3 (q, br), 130.7, 129.7, 129.5, 128.8,
1744
᎐
aromatic)
128.5, 128.4, 128.2, 126.9, (overlap of one aromatic CH
signal), 63.9 (C-7), 51.9 (OCH3)d
3.44 (3 H, s, OCH3), 3.86 (3 H, s, ArOCH3), 4.31 (1 H, d, 172.1 (C᎐O, br), 158.9 (q), 142.2 (q), 141.3 (q), 132.6 (q), 1733
᎐
J 11.6, 7-Hax), 4.47 (1 H, d, 7-Heq), 5.18 (1 H, s, 5-H), 7.0– (remaining quaterenary carbons were not seen due to being
7.62 (11 H, m, aromatic)
too broad), 130.5, 129.9, 129.4, 128.6, 128.4, 128.1, 126.3,
113.6, 64.2 (C-7), 55.1 (OCH3), 51.9 (OCH3)d
15m
3.42 (3 H, s, OCH3), 4.25 (1 H, d, J 11.9, 7-Hax), 4.44 (1
H, d, J 11.9, 7-Heq), 5.20 (1 H, s, 5-H), 7.25–7.84 (11 H,
m, aromatic)
a Compounds identified in Table 1. b At 328 K and in CDCl3 unless otherwise stated. c At 298 K and in CDCl3 unless otherwise stated. d C-5 peak
broad and obscured by CDCl3 peaks.
benzaldehyde 18 (1.8 g, 6.9 mmol), methyl chloroacetate (2.6 g,
24.1 mmol) and sodium methoxide (1.11 g, 20.55 mmol) by the
method given above for 9a, with a reaction time of 6 h. The
product 20 was crystallised from hexane–ether (80:20) as a
white solid (1.75 g, 76%), mp 107–108 ЊC, from hexane–ether
(90:10) (Found: C, 57.8; H, 3.95%; Mϩ, 332.0036. C16H13BrO3
requires C, 57.65; H, 3.9%; M, 332.0048); δH 3.0 (1 H, d, J 2.0,
2-H), 3.67 (3 H, s, OCH3), 4.32 (1 H, d, J 2.0, 3-H) and 7.29–
as a white solid (0.16 g, 64%) mp 141–142 ЊC (Found: C,
57.4; H, 3.6%; Mϩ, 332.0057. C16H13BrO3 requires C, 57.65;
H, 3.9%; M, 332.0057); δH(328 K) 3.48 (3 H, s, CH3), 4.62 (1
H, d, J 12, 7-Hax), 4.76 (1 H, d, J 12, 7-Heq), 5.05 (1 H, s, 5-
H), 7.23–7.66 (7 H, m, aromatic); δC 171.3 (C᎐O), 142.6 (q),
᎐
139.6 (q), 133.8 (q), 133.1 (q), 125.1 (q), 132.5, 130.0, 129.5,
128.9, 127.8, 126.9 (CH), (overlap of one CH signal), 76.6
(C-5), 65.9 (C-7) and 52.0 (CH3); νmax(Nujol)/cmϪ1 1723
7.61 (8 H, m, aromatic); δ 168.4 (C᎐O), 144.0 (q), 139.6 (q),
(C᎐O).
᎐
᎐
C
132.1 (q), 124.7 (q), 131.6, 129.5, 129.2, 128.4, 128.2, 127.5
(E)-5-Methoxycarbonyl-8-(phenylethenyl)-5,7-dihydrodibenz-
(CH), 57.9 (C-3), 55.8 (C-2) and 52.2 (OCH3); νmax(Nujol)/cmϪ1
[c,e]oxepine 15b. 4-Bromo-7-methoxycarbonyl-5,7-dihydrodi-
benz[c,e]oxepine 23 (0.25 g, 0.75 mmol) and (E)-2-phenyl-
ethenylboronic acid8 (0.32 g, 2.16 mmol) were coupled using
the method given above for 19a, with a reaction time of
24 h. The product 15b (0.23 g, 86%) was crystallised from
hexane.
1730 (C᎐O).
᎐
4-Bromo-7-methoxycarbonyl-5,7-dihydrodibenz[c,e]oxepine
23. The oxirane 20 (0.25 g, 0.75 mmol) on FVP (distillation
temperature 130–160 ЊC, time 1 h) followed by dry flash chro-
matography (silica, hexane–ether, 90:10) gave compound 23
3032
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
Table 8 Product ratios from the pyrolysis of the oxiranes 9b–m and 10f,g
Oxiranes 9
Oxiranes 10
Nitrile ylide
product ratio 5:42
Substituent
Prod. ratio 13:15
Yield (%)
Prod. ratio 14:16
Yield (%)
b (E)-2-Phenylethenyl
c 2-Thienyl
20
9
8
1.5
0.7
0.8
65
70
75
75
70
70
70
70
75
70
75
80
>100
>100
d 3-Thienyl
e 3,5-Dichlorophenyl
f 3,5-Bis(trifluoromethyl)phenyl
g 3,5-Dimethyphenyl
h 3-Nitrophenyl
(f) 0.7
(g) 0.9
75
75
32.0
8.3
>100
1.3 (2Ј/6Ј = 3.7)
i
j
4-Fluorophenyl
4-Methylphenyl
1.2
1.3
1.2
1.2
1.3
1.2
1.5
2.2
1.6
2.8
k 4-Chlorophenyl
4-Methoxyphenyl
m 4-Trifluoromethylphenyl
l
Table 9 1H NMR spectral data on the products of the pyrolysis of the
oxiranes 9b–m and 10f,g
nitrogen. The solution was stirred at room temperature for 2 h
and worked up as for 22f to give the product as a yellow oil
(0.4 g, 79%) (HRMS: found Mϩ, 360.1878. C28H24 requires M,
360.1871); δH 2.39 (s, 6 H, 2 × CH3), 6.06 (1 H, d, J 16.7, ethenyl
CH), 6.88 (1 H, d, J 16.7, ethenyl CH), 7.02–7.47 (16 H, m,
aromatic).
1H NMR data [δ (J Hz)] at 328 K
Reactant
Product(s)
5-H (s)
7-H, 7Ј-H
9b
9c
9d
9e
9f
15b
13b
15c
13c
15d
13d
15e
13e
15f
5.09
5.37
5.17
5.48
5.17
5.48
5.18
6.00
5.24
5.80
5.22
5.68
5.22
5.20
4.62 (br)
2-Phenyl-3-{2-phenyl-6-[3,5-bis(trifluoromethyl)phenyl]-
phenyl}oxirane 10f. A solution of (E)-1-phenyl-2-{2-phenyl-6-
[3,5-bis(trifluoromethyl)phenyl]}ethene 22f (0.5 g, 1.06 mmol)
and m-chloroperbenzoic acid (80%; 0.7 g, 3.24 mmol) in dichlo-
romethane was stirred at room temperature for 48 h. The solu-
tion was washed with aqueous sodium carbonate (5% w/v;
2 × 30 cm3), dried and the solvent was evaporated to give the
product as a colourless oil which was crystallised to give a white
solid (0.3 g, 82%, mp 57–58 ЊC from hexane) (Found: C, 69.35;
H, 3.75%; Mϩ, 484.1266. C28H18F6O requires C, 69.4; H, 3.7%;
M, 484.1262); δH 3.15 (1 H, d, J 2.1, 2-H), 4.10 (1 H, d, J
2.1, 3-H), 6.51–6.56 (2 H, m, aromatic), 7.11–7.57 (10 H, m,
aromatic) and 8.0–8.08 (4 H, m, aromatic); δC 143.4 (q), 143.3
(q), 140.4 (q), 140.3 (q), 135.8 (q), 132.8 (q), 131.2 (q), 130.2,
129.4, 129.0, 128.9, 128.5, 128.4, 127.9, 127.8, 127.7, 127.4,
4.65, 4.85 (d, J 12.5)
4.40, 4.51 (d, J 11.5)
4.60, 4.71 (d, J 12.9)
4.38 (br), 4.55 (br)
superimposed on 15d
4.23, 4.45 (d, J 11.9)
4.03, 4.73 (d, J 12.9)
4.20, 4.50 (d, J 12.0)
4.00, 4.82 (d, J 12.0)
4.35, 4.50 (d, J 11.6)
4.10, 4.73 (d, J 11.6)
4.28, 4.45 (d, J 11.7)
4.23, 4.56 (d, J 12.0)
13f
9g
9h
15g
13g
15h
(2-NO2)
13h
(4-NO2)
15i
5.26
5.19
5.14
5.20
5.17
5.19
5.15
5.20
5.17
5.21
5.25
5.55
6.55
5.63
6.37
obscured
9i
4.27, 4.48 (d, J 11.8)
4.31, 4.46 (d, J 11.6)
4.33, 4.49 (d, J 11.7)
superimposed on 15j
4.27, 4.47 (d, J 11.7)
4.28, 4.52 (d, J 11.6)
4.31, 4.47 (d, J 11.6)
superimposed on 15l
4.26, 4.46 (d, J 11.9)
3.94, 4.56 (d, J 12.0)
4.16, 4.62 (d, J 11.9)
4.17, 5.02 (d, J 12.0)
4.21, 4.87 (d, J 11.5)
4.34, 4.86 (d, J 12.0)
13i
1
124.7, 120.9 (CF3), 61.6 (C-3) and 61.4 (C-2). The H NMR
9j
15j
spectrum also showed the presence of a small quantity of a
second isomer with absorptions at δH 4.52 (1 H, d, J 2.5, 2-H)
and 4.87 (1 H, d, J 2.5, 3-H).
13j
9k
9l
15k
13k
15l
2-Phenyl-3-[2-phenyl-6-(3,5-dimethylphenyl)phenyl]oxirane
10g. A solution of (E)-1-phenyl-2-[2-phenyl-6-(3,5-dimethyl-
phenyl)phenyl]ethene 22g (0.35 g, 0.97 mmol) and m-chloro-
perbenzoic acid (80%; 0.34 g, 1.67 mmol) in dry dichloro-
methane (30 cm3) was stirred at room temperature for 48 h. The
solution was washed with aqueous sodium carbonate (5% w/v;
2 × 30 cm3), dried and the solvent was evaporated to give the
product as a yellow oil (0.3 g, 82%) (HRMS: found Mϩ,
376.1830. C28H24O requires M, 376.1827); δH 2.3 (6 H, s,
2 × CH3), 3.21 (1 H, d, J 2.1, 2-H), 4.18 (1 H, d, J 2.1, 3-H),
6.57–6.62 (2 H, m, aromatic) and 7.1–7.56 (14 H, m, aromatic);
δC 143.4 (q), 143.3 (q), 141.4 (q), 141.0 (q), 137.7 (q), 136.6 (q),
132.4 (q), 129.0, 128.9, 128.6, 128.2, 127.9, 127.5, 127.0, 126.4,
125.0, 62.3 (C-3), 61.2 (C-2) and 21.1 (CH3).
13l
9m
10f
10g
15m
13m
16f
14f
16g
14g
Preparation of 2-phenyl-3-[2-phenyl-6-(substituted)phenyl]-
oxiranes 10f,g
These were prepared by the route shown in Scheme 2.
(E)-1-Phenyl-2-{2-phenyl-6-[3,5-bis(trifluoromethyl)phenyl]}-
ethene 22f. 2-Phenyl-6-[3,5-bis(trifluoromethyl)phenyl]benz-
aldehyde 19f (0.8 g, 2.03 mmol), in DME (10 cm3) was added
over 1 h to a solution of diethyl benzylphosphonate (0.55 g,
2.41 mmol) and sodium hydride (0.1 g; 60% in oil). The solution
was stirred at room temp. for 12 h. Evaporation of the solvent
and dry flash chromatography (silica, hexane–ether, 70:30) of
the residue gave the product (0.93 g, 97%) as a colourless oil
(HRMS: found Mϩ, 468.1311. C28H18F6 requires M, 468.1313);
HPLC analysis showed the E:Z ratio to be >98:1; δH 6.0 (1 H,
d, J 16.3, ethenyl CH), 6.78 (1 H, d, J 16.3, ethenyl CH) and
6.9–7.9 (16 H, m, aromatic).
Preparation of the 5-phenyl-8-substituted-5,7-dihydrodibenz-
[c,e]oxepines 16f,g
These compounds were prepared by the route shown in Scheme 2.
(E)-2-Bromo-6-phenylstilbene. 2-Bromo-6-phenylbenzalde-
hyde 18 (1.0 g, 3.86 mmol), in DMF (5 cm3) was added over 10
min to a solution of diethyl benzylphosphonate (1.22 g, 5.34
mmol) and sodium methoxide (0.29 g, 5.37 mmol) in DMF (50
cm3). The solution was stirred at room temperature for 2 h.
Evaporation of the solvent, dry flash chromatography (silica,
hexane–ether, 70:30) of the residue and crystallisation from
hexane gave the product (1.1 g, 85%) as a white solid, mp 74.5–
75.5 ЊC (HRMS: found Mϩ, 334.0364, 336.0333. C20H15Br
requires M, 334.0357, 336.0338); δH 6.37 (1 H, d, J 16.5), 7.02 (1
(E)-1-Phenyl-2-[2-phenyl-6-(3,5-dimethylphenyl)phenyl]-
ethene 22g. 2-Phenyl-6-(3,5-dimethylphenyl)benzaldehyde 19g
(0.4 g, 1.4 mmol), in DMF (10 cm3) was added over 10 min to a
solution of diethyl benzylphosphonate (0.35 g, 1.53 mmol) and
sodium ethoxide (0.1 g, 1.47 mmol) in DMF (15 cm3), under
J. Chem. Soc., Perkin Trans. 1, 1997
3033

View Article Online
H, d, J 16.5), 7.2–7.4 (12 H, m, aromatic) and 7.63 (1 H, dd, J
8.1, 1.2, aromatic).
stirred for 20 min. A solution of 2-formylphenylboronic acid
(0.5 g, 3.3 mmol) and sodium carbonate (0.35 g, 3.3 mmol) in
water (5 cm3) was added and the mixture was heated under
reflux for 12 h. The usual work-up followed by Kugelrohr distil-
lation gave the product (0.44 g, 66%) as a colourless oil,3 Kugel-
rohr oven temperature 170–180 ЊC at 0.05 mmHg; shown by
mass spectroscopy to be 98% monodeuteriated; δH 7.19–7.58 (7
H, m), 7.96 (1 H, dd, J 7.7, 1.1) and 9.92 (1 H, s, CHO).
2-Phenyl-3-[(2-bromo-6-phenyl)phenyl]oxirane 21. Reaction
of (E)-2-bromo-6-phenylstilbene (2.5 g, 7.46 mmol) and m-
chloroperbenzoic acid (80%; 6.4 g, 11.92 mmol) using the
method described for 22f gave 21 as a colourless oil (0.3 g,
82%) bp 180–190 ЊC/0.5 mmHg (Found: C, 68.15; H, 4.45%;
Mϩ, 350.0336. C20H15BrO requires C, 68.4; H, 4.25%; M,
350.0306); δH 3.40 (1 H, d, J 2.1, 2-H), 4.05 (1 H, d, J 2.1, 3-
H), 4.57 (1 H, d, J 2.3, 2-H), 5.47 (1 H, d, J 2.3, 3-H),
(mixture of isomers, ratio ca. 1:5 respectively) and 6.89–7.93
(13 H, m, aromatic).
4-Bromo-7-phenyl-5,7-dihydrodibenz[c,e]oxepine 24. The oxir-
ane 21 (0.25 g, 0.71 mmol) on FVP (distillation temperature
150–180 ЊC, time 1 h) followed by dry flash chromatography
(silica, hexane–ether, 90:10) gave compound 24 (0.15 g, 75%) as
a white solid, mp 141–142 ЊC (Found: C, 68.5; H, 4.6%; Mϩ,
350.0303. C20H15BrO requires C, 68.4; H, 4.3%; M, 350.0306);
δH(297 K) 4.33 (1 H, d, J 11.7, 7-Hax), 5.21 (1 H, d, J 11.7,
7-Heq), 5.39 (1 H, s, 5-H), 6.86 (1 H, m, aromatic), 7.28–7.53 (10
H, m, aromatic) and 7.64–7.67 (1 H, m, aromatic); δC 143.3 (q),
140.2 (q), 139.6 (q), 137.8 (q), 133.9 (q), 124.6 (q), 132.3, 129.7,
129.4, 128.7, 128.4, 128.3, 127.9, 127.5, 127.3, 126.8 (CH), 77.1
(C-5) and 66.3 (C-7).
5-Phenyl-8-[3,5-bis(trifluoromethyl)phenyl]-5,7-dihydrodi-
benz[c,e]oxepine 16f. 4-Bromo-7-phenyl-5,7-dihydrodibenz-
[c,e]oxepine 24 (0.1 g, 0.28 mmol) and 3,5-bis(trifluoro-
methyl)benzeneboronic acid (0.1 g, 0.36 mmol) were coupled
using the method described for 19a with a reaction time of 10 h
to give 16f (0.11 g, 80%) as an oil (HRMS: found Mϩ,
484.12602. C28H18F6O requires M, 484.12618); δH(297 K) 4.18
(1 H, d, J 11.9, 7-Hax), 4.64 (1 H, d, J 11.9, 7-Heq), 5.55 (1 H, s,
5-H), 6.89 (1 H, d, aromatic) and 7.34–8.04 (14 H, m, aromatic);
δC 142.7 (q), 142.3 (q), 140.6 (q), 139.7 (q), 139.2 (q), 137.9 (q),
132.3 (q), 131.2 (q), 129.7, 129.5, 129.4, 129.1, 129.0, 128.8,
128.4, 128.0, 127.9, 127.7, 127.3, 127.2 (CH), 121.1 (CF3), 77.1
(C-5) and 63.7 (C-7).
5-Phenyl-8-(3,5-dimethylphenyl)-5,7-dihydrodibenz[c,e]-
oxepine 16g. 4-Bromo-7-phenyl-5,7-dihydrodibenz[c,e]oxepine
24 (0.1 g, 0.28 mmol) and 3,5-dimethylphenylboronic acid (0.08
g, 0.56 mmol) were coupled using the method described for 19a
with a reaction time of 10 h to give 16g (0.09 g, 85%) as an oil
(HRMS: found Mϩ, 376.1835. C28H24O requires M, 376.1827);
δH(297 K) 2.42 (6 H, s, 2 × CH3), 4.21 (1 H, d, J 11.5, 7-Hax),
4.87 (1 H, d, J 11.5, 7-Heq), 5.63 (1 H, s, 5-H) and 6.64–7.74 (15
H, m, aromatic).
2-Methoxycarbonyl-3-(2-[2-2H]phenylphenyl)oxirane. Sodium
methoxide (0.47 g, 8.6 mmol) was added to a solution of 2-[2-
2H]phenylbenzaldehyde (0.52 g, 2.84 mmol) and methyl chlo-
roacetate (0.99 g, 9.1 mmol) in THF (25 cm3) and the mixture
was heated under reflux for 5 h. The usual work-up followed by
Kugelrohr distillation gave the product (0.55 g, 76%) as a pale
yellow oil,3 Kugelrohr oven temperature 190–220 ЊC at 0.01
mmHg; shown by mass spectroscopy to be 98% monodeuteri-
ated (HRMS: found Mϩ, 255.1001. C16H132HO3 requires M,
255.1006); δH 3.44 (1 H, br s, 3-H), 3.69 (3 H, s, CH3), 3.92 (1 H,
br s, 2-H) and 7.2–7.4 (8 H, m, aromatic); m/z (EI) 255 (9%),
196 (28), 195 (23), 183 (34), 182 (47), 181 (12), 168 (21), 167
(61), 166 (100), 165 (38), 154 (17), 153 (27), 152 (13) and 77 (10).
FVP of 2-Methoxycarbonyl-3-(2-[2-2H]phenylphenyl)oxirane.
Using the general FVP method, the reactant (0.044 g) was dis-
tilled into the furnace tube using an inlet temperature of 120–
140 ЊC. Dry flash chromatography (silica, ether–hexane, 3:1) of
the product gave a mixture of 5-methoxycarbonyl-5,7-dihydro-
[1-2H]- and -[7-2H]-dibenz[c,e]oxepines 35 and 37 (0.017 g, 39%)
shown by mass spectrometry to be 98% monodeuteriated
(HRMS: found Mϩ, 255.1008. C16H132HO3 requires M,
255.1006); δH 3.43 (3 H, s, CH3), 4.45 (ca. 1.5 H, br s, 7-H), 4.99
(1 H, s, 5-H) and 7.2–7.5 (ca. 7.5 H, m, aromatic); δ2 (CHCl3)
H
4.47 (br) and 7.58 (br) integral ratio (1:1.02); m/z (EI) 255
(21%), 197 (22), 196 (100), 195 (22), 169 (17), 168 (85), 167 (28),
166 (55), 165 (22), 153 (18), 152 (11) and 83 (11). Two further
2
experiments gave yields of 38 and 35% and H NMR integral
ratios of 0.99 and 0.97 (average = 0.99 0.02).
Non-deuteriated samples of 2-methoxycarbonyl-3-(2-
phenylphenyl)oxirane3 and 5-methoxycarbonyl-5,7-dihydro-
dibenz[c,e]oxepine3 were prepared by the same routes as the
deuteriated samples for comparison of their NMR and mass
spectra.
Acknowledgements
We thank the ESPRC for a postdoctoral fellowship (D. F. O’S)
and Dr H. McNab for advice on flash vacuum pyrolysis.
Intramolecular competition reactions of the carbonyl ylides 11b–
m and 12f,g via FVP of the oxiranes 9b–m and 10f,g
References
General method. Using the general FVP method specified
earlier, the reactant (ca. 0.2 g) was distilled into the furnace tube
(625 ЊC) over 1 h using an inlet temperature of 150–180 ЊC. The
ratio of the cyclised products (13:15 and 14:16, Table 8), was
measured from the crude product mixture by two methods; (i)
1H NMR spectroscopy (200 or 360 MHz) at 328 K via com-
parative integration of the 5-H absorptions, and (ii) HPLC
using a 5 µm silica column (250 × 4.6 mm), eluting with hexane
for the products from 10f,g, or with hexane–ether (80:20 to
90:10 as appropriate) for the products from 9b–m. One of the
products in each case was identified by comparison with the
appropriate ‘authentic’ sample, 15b–m or 16f,g via peak
enhancement where appropriate. The products were then separ-
ated out as a mixture by dry column flash chromatography
(silica, hexane–ether, 60:40 to 80:20 as appropriate), the yields
are given in Table 8 and their 1H NMR spectra in Table 9.
1 (a) G. Zecchi, Synthesis, 1991, 181; (b) K. R. Motion, I. R.
Robertson, J. T. Sharp and M. D. Walkinshaw, J. Chem. Soc., Perkin
Trans. 1, 1992, 1709; (c) P. W. Groundwater and J. T. Sharp,
Tetrahedron, 1992, 48, 7951; (d ) K. E. Cullen and J. T. Sharp, J. Chem.
Soc., Perkin Trans. 1, 1993, 2961; (e) H. Finch, D. H. Reece and
J. T. Sharp, J. Chem. Soc., Perkin Trans. 1, 1994, 1193; ( f ) A. J. Blake,
M. Harding and J. T. Sharp, J. Chem. Soc., Perkin Trans. 1, 1994,
3149.
2 K. E. Cullen and J. T. Sharp, J. Chem. Soc., Perkin Trans. 1, 1995,
2565.
3 W. Eberbach and U. Trostmann, Chem. Ber., 1985, 4035.
4 D. F. O’Shea and J. T. Sharp, J. Chem. Soc., Perkin Trans. 1, 1996,
515.
5 H. B. Hass and M. L. Bender, J. Am. Chem. Soc., 1949, 71, 1767.
6 T. K. Miller, J. T. Sharp, H. R. Sood and Edward Stefaniuk, J. Chem.
Soc., Perkin Trans. 2, 1984, 823.
7 J. T. Sharp, I. Gosney and A. G. Rowley, Practical Organic
Chemistry, a Student Handbook of Techniques, Chapman and Hall,
London, 1989, p. 160.
Preparation and FVP of 2-methoxycarbonyl-3-(2-[2-2H]phenyl-
phenyl)oxirane
8 H. C. Brown and S. K. Gupta, J. Am. Chem. Soc., 1975, 97, 5249.
2-[2-2H]Phenylbenzaldehyde. Bromo[2-2H]benzene6 (0.58 g,
3.68 mmol) and tetrakis(triphenylphosphine)palladium (0.11 g,
0.095 mmol) in 1,2-dimethoxyethane (DME) (20 cm3) were
Paper 7/02962D
Received 30th April 1997
Accepted 2nd June 1997
3034
J. Chem. Soc., Perkin Trans. 1, 1997