Syntheses, structure analyses, and reactions of 1,3,5-trioxepanes and related compounds

Article abstract of DOI:10.1039/a802928h

Acid-catalysed condensations of 1,5- or 1,6-dicarbonyl compounds with ethylene glycol give 1,3,5-trioxepane derivatives as a result of neighbouring participation by the adjacent carbonyl group during the acetalization process. With trimethylene glycol, the related 1,3,5-trioxocanes have also been obtained. Reaction of the 1,3,5-trioxepanes with (a) Grignard reagents gives dialkyl-substituted cyclic ethers, (b) titanium tetrachloride-allyltributyltin gives diallyl-substituted cyclic ethers and (c) triethylsilane in the presence of trimethylsilyl triflate provides the corresponding cyclic ethers.

Full text of DOI:10.1039/a802928h

View Article Online / Journal Homepage / Table of Contents for this issue
Syntheses, structure analyses, and reactions of 1,3,5-trioxepanes and
related compounds
Kevin J. McCullough,*,a Araki Masuyama,b Keith M. Morgan,a
Masatomo Nojima,*,b Yuji Okada,b Syuzo Satakeb and
Shin-ya Takedab
a Department of Chemistry, Heriot-Watt University, Edinburgh EH14 4AS,
Scotland
b Department of Materials Chemistry, Faculty of Engineering, Osaka University,
Suita, Osaka 565, Japan
Acid-catalysed condensations of 1,5- or 1,6-dicarbonyl compounds with ethylene glycol give 1,3,5-
trioxepane derivatives as a result of neighbouring participation by the adjacent carbonyl group during
the acetalization process. With trimethylene glycol, the related 1,3,5-trioxocanes have also been
obtained. Reaction of the 1,3,5-trioxepanes with (a) Grignard reagents gives dialkyl-substituted
cyclic ethers, (b) titanium tetrachloride–allyltributyltin gives diallyl-substituted cyclic ethers and
(c) triethylsilane in the presence of trimethylsilyl triflate provides the corresponding cyclic ethers.
Recent increased use of acetals in cross-coupling reactions with
nucleophiles has led to the development of a variety of methods
Results and discussion
for their activation.1 In this respect, the acetalization of di-
carbonyl compounds, in which the two carbonyl groups lie in
close proximity to each other, may proceed by a neighbouring
group participation affording bicyclic compounds bearing a
‘bis-acetal ether’ functionality.2,3 Since such compounds may
behave as potent dication equivalents, they should, in principle,
be able to couple with two nucleophiles (Scheme 1) and hence
provide new methods for the synthesis of cyclic ethers.4
Acetalization of dicarbonyl compounds
To investigate the possibility of neighbouring group partici-
pation during the acetalization of dicarbonyl compounds, a
mixture of the keto aldehyde 1 and ethylene glycol was treated
with chlorosulfonic acid in dichloromethane (Scheme 2). By
column chromatography on silica gel, the expected 1,3,5-
trioxepane 2 was isolated in high yield (78%). The analogous
reaction between 1 and trimethylene glycol afforded the corre-
sponding 1,3,5-trioxocane 3 along with a roughly similar
amount of the keto acetal 4. With 2,2-dimethylpropane-1,3-
diol, however, only the keto acetal 5 was isolated (81%) because
the isomeric 1,3,5-trioxocane 6 was not produced in significant
quantities. The corresponding condensation reactions involving
the structurally rigid 1,6-dialdehyde 7 yielded an analogous
range of products 8–10 (Scheme 3).
H+
+
O
O
O
O
O
+
O
HOCH2CH2OH
Scheme 1
O
Ph
O
O
O
O
O
Ph
O
O
H+
H+
Ph
O
O
+
Ph
O
HO
OH
HO
OH
2 (78%)
4 (30%)
1
3 (34%)
H+
HO
OH
O
O
O
O
Ph
Ph
O
O
6 (0%)
5 (81%)
Scheme 2
J. Chem. Soc., Perkin Trans. 1, 1998
2353

View Article Online
membered 1,3,5-trioxepanes. In the condensation reaction
between keto aldehyde 1 and 2,2-dimethylpropane-1,3-diol,
the preponderant formation of the 1,3-dioxane derivative 5
possibly reflects the steric effect of the geminal methyl groups
which may be more readily accommodated in the six-membered
ring than the alternative eight-membered ring.
O
O
O
In a series of acid-catalyzed condensation reactions between
1,5-dicarbonyl compounds 12 and ethylene glycol, the keto
aldehydes 12c–e also gave the corresponding trioxepanes 13c–e
in good yield (Table 1). In contrast, however, only the isomeric
keto acetals 14a,b were obtained from keto aldehydes 12a,b.
Although these results are remarkably clear cut and suggest
that the course of the acetalization was influenced by the
structures of the dicarbonyl compounds, there is no obvious
correlation between the structure of the substrate and that of
the product.
The analogous reactions involving trimethylene glycol
give variable results; keto aldehyde 12c yields the trioxocane
derivative 15 whereas 12a and 12d are transformed into the
mono acetals 16 and 17 respectively. This latter result is
surprising given that the reaction of the keto aldehyde 12d
with ethylene glycol resulted in exclusive formation of the tri-
oxepane derivative 13d. These observations tend to reinforce
the view that the neighbouring group participation and cycliz-
ation processes leading to the eight-membered trioxocane
appear to be less efficient than those related to the formation
of the seven-membered trioxepane.
9 (17%)
H+
HO
OH
O
O
O
O
H+
O
HO
OH
8 (91%)
7
H+
HO
OH
Reactions of the dicarbonyl compounds 1, 6 and 12c with
methanol under acidic conditions gave the corresponding
bis-acetal ethers 18, 19 and 20 respectively, albeit in moderate
yields.
O
O
O
Structural studies of 1,3,5-trioxepanes and related compounds
The structures of the bicyclic acetals 2 and 3, and the keto
acetal 4 in the solid state were determined by X-ray crystal-
lographic analysis and are depicted in Figs. 1–3 respectively.
Although the observed geometrical parameters for these
molecules are unexceptional and lie well within expected
ranges, further investigation of their respective overall
structures, in particular their preferred conformations, and
comparison with structures generated by molecular modelling
could offer some rationale for the product selectivities
mentioned above.
Molecular modelling of the structure of the bicyclic acetal
2 using molecular mechanics (MM2) combined with a Monte
Carlo conformational search procedure reveals that while most
of the molecule is rigid and essentially planar, the 1,3,5-
trioxepane ring is flexible, being able to adopt several distorted
O
10 (61%)
Scheme 3
The stepwise mechanism outlined in Scheme 4 tentatively
accounts for the formation of the trioxepane derivative 2
from the keto aldehyde 1. Since the 1,3-dioxolane 11 was not
formed in significant quantities, it is presumed that the stabi-
lized carbocation Y rather than X must be the key intermediate
in this reaction pathway. When ethylene glycol is replaced by
trimethylene glycol, however, there is clearly a finer balance
between the analogous competing pathways consistent with
the notion that the eight-membered 1,3,5-trioxocane would be
less readily formed by cyclization than the corresponding seven-
OH
OH
a
O
OH
O
O
O
Ph
O
O
/ H+
+H+, – H2O
Ph
– H+
O
+
Ph
HO
OH
b
1
X
a
X
11
+H+, – H+
b
OH
O
OH
OH
Ph
O
O
O
Ph
– H+
+H+, – H2O
+
2
Y
Scheme 4
2354
J. Chem. Soc., Perkin Trans. 1, 1998

View Article Online
Table 1 Isolated products from the acid-catalysed condensation reaction between dicarbonyl compound 12 and ethylene glycol
R1
R1
R1
O
O
O
H+
O
+
O
O
O
R2
R4
O
R4
R2
HO
OH
R2
R3
R3
R3
R4
12
13
14
Isolated yield (%)
Compound 12
R1
R2
R3
R4
13
14
a
b
c
d
e
Ph
Ph
Ph
H
H
H
H
Me
H
H
H
H
Ph
Ph
—
—
55
47
95
76
83
—
—
—
Me
Me
Ph
H
᎐(CH2)4᎐
Ph
O
Ph
O
O
O
O
O
O
O
Me
Me
Ph
O
Ph
17 (78%)
15 (60%)
16 (69%)
MeO
OMe
OMe
Ph
OMe
O
Ph
O
O
OMe
H
OMe
Me Me
20
19 (64%)
18 (41%)
Scheme 5
Fig. 2 The solid state structure of one molecule of 1,3,5-trioxocane 3
(ORTEP,15 the non-hydrogen atoms are represented by 50% probability
ellipsoids and hydrogen atoms by spheres of arbitrary radius)
Fig. 1 The solid state structure of one molecule of 1,3,5-trioxepane 2
(ORTEP,15 the non-hydrogen atoms are represented by 50% probability
ellipsoids and hydrogen atoms by spheres of arbitrary radius)
Fig. 3 The solid state structure of one molecule of keto acetal 4
(ORTEP,15 the non-hydrogen atoms are represented by 50% probability
ellipsoids and hydrogen atoms by spheres of arbitrary radius)
J. Chem. Soc., Perkin Trans. 1, 1998
2355

View Article Online
Table 2 Summary of the molecular modelling results
Conformation of 1,3,5-
Strain energy (MM2)/
kcal molϪ1
∆Hf (MOPAC-PM3)/
Compound
trioxepane/trioxocane ring
kcal molϪ1
2
chair
chair
chair
boat-chair
twist-chair-chair
crown
—
—
17.764
17.817
18.584
18.058
22.000
22.691
17.701
18.639
19.445
18.518
Ϫ34.454
Ϫ35.725
Ϫ35.006
Ϫ37.678
Ϫ37.625
Ϫ39.473
Ϫ35.902
Ϫ47.445
Ϫ47.232
Ϫ32.214
3
4
5
6
boat-chair
—
11
chair conformations with shallow minima. The two lowest
energy conformations were found to have similar strain energies
and heats of formation and a third, having a 1,3,5-trioxepane
ring with mirror plane symmetry, was less favoured than the
other two by about 1 kcal molϪ1 (Table 2). The lowest energy
structure calculated for 2 is comparable to that determined
by X-ray crystallography (Fig. 1). The heat of formation
(MOPAC93-PM3) for the unobserved keto acetal 11 was
estimated to be about 2 kcal molϪ1 higher than that of 2.
I
δ
Mg
I
O
Mg
O
O
O
O
O
Me
Me
Ph
δ
MeMgI
Ph
Ph
O
O
O
2
MeMgI
5%
Structural modelling of the bicyclic acetal 3 indicates that the
1,3,5-trioxocane ring adopts three well-defined, recognizable
conformations: a boat-chair, a twist-chair-chair and a crown.
The structure of molecule 3 in the solid state (Fig. 2) corre-
sponds to a boat-chair conformation, essentially the same as
the global minimum identified by molecular mechanics.
H
Me
Ph
O
δ
Mg
I
Me
OMgI
a
O
Me
b
δ
Me
Ph
H
Cursory inspection of the calculated and solid state molecu-
lar structures of keto acetal 4 reveals marked differences in the
orientations of the 1,3-dioxanyl and benzoyl groups. Since the
substituents in keto acetal 4 would be expected to have a degree
of conformational freedom, the observed conformation of 4
in the solid state will probably be determined by crystal packing
forces. From heats of formation calculated for minimised
structures, the bicyclic acetal 3 is more stable than the keto
O
c-21 (75%)
Me
Mg
a
Me
H
O
Me
Ph
I
b
t-21 (10%)
acetal 4 by about 1.7 kcal molϪ1
.
Structural modelling of the mono-acetal 5 and the isomeric
gem-dimethyl 1,3,5-trioxocane derivative 6 indicate that the
strain energy or heats of formation of these molecules differ by
less than 1 kcal molϪ1 which would not readily account for the
high degree of selectivity observed in the reaction.
O
Ph
Ph
Me
O
0%
MeMgI
O
H
O
OH
O
Me
Me
Me
Me
In summary, molecular modelling studies using molecular
mechanics (MM2) have accurately reproduced the preferred
conformations of the bicyclic acetals 2 and 3 observed in the
solid state. The heats of formation estimated by semi-empirical
calculations are less reliable in this respect and the differences
in heats of formation between the pairs of isomeric mono-
and bi-cyclic acetals do not correlate particularly well with the
observed product distributions.
22 (52%)
13c
Ph
Ph
4%
OMe
Me
O
4%
MeMgI
O
H
H
OMe
OMe
Me
Me
20
Me
Me
23 (66%)
Reaction of 1,3,5-trioxepanes
Reaction with Grignard reagents. Acetals are reported to react
with Grignard reagents to provide the alkyl-substituted ethers.5
By analogy, the reaction of the trioxepane 2 with methyl-
magnesium iodide afforded the dimethyl-substituted cyclic
ether as a mixture of diastereoisomers c-21 and t-21.† On the
basis of 1H NMR NOE studies, the major product was found to
be stereoisomer c-21 (Scheme 5). In contrast, the reaction of the
trioxepane 13c with methylmagnesium iodide gave only the
mono-methylated compound 22; similarly the bis-acetal 20 was
transformed into 23. These latter results imply that the mag-
nesium ion of the Grignard reagent must selectively coordinate
Scheme 5
to the more hindered oxygen atom in 13c and 20 resulting
in cleavage of the C᎐O bond with concomitant development
of a significant degree of electrophilic character at the highly
substituted carbon centre. The nucleophilic methyl group
would then be subsequently delivered to the incipient carbo-
cationic centre. 1H NMR NOE Measurements on 22 and 23 are
consistent with the methyl group having been delivered syn to
the displaced oxygen. Further methylation of either 22 or 23
does not occur because the required carbocation, formed on
heterolytic cleavage of the C᎐O bond, cannot be sufficiently
stabilised. The mono-methylated product from 2 does, however,
undergo a second methylation predominantly via an inter-
molecular displacement to yield c-21 as illustrated in Scheme 5.
† Nomenclature based on the fiducial substituent system proposed by
L. C. Cross and W. Klyne, Pure Appl. Chem., 1976, 45, 11; see also E. L.
Eliel and S. H. Wilen in ‘Stereochemistry of Organic Compounds’,
Wiley, New York, 1994, pp. 665–666.
2356
J. Chem. Soc., Perkin Trans. 1, 1998

View Article Online
Table 3 Products from the reaction of Grignard reagents with com-
pound 2
5%
O
O
H
O
Ph
CH2=CHCH2SnBu3
TiCl4
Ph
O
O
O
R
H
R
R
O
O
Ph
Ph
H
RMgBr
R
Ph
O
+
c-24d (49%)
2
Ph
c-24
t-24
2
8%
O
Product
R
Yield (%)
c:t
H
24a
24b
24c
24d
Et
56
28
34
46
86:14
100:0
n.a.
Pri
Ph
c-26
allyl
100:0
HO
O
O
7%
H
CH2=CHCH2SnBu3
TiCl4
H
O
O
Pri
Ph
Ph
Ph
12e
27
Scheme 6
c-24b
1
Table 4 Products obtained from the reaction between Grignard
reagents and 1,3,5-trioxepane 8
ponding trioxepanes 2, 8 and 13c respectively (Scheme 6). H
NOE NMR Measurements suggest that compounds 24d and
26 were each formed stereoselectively as the c-isomer in which
the allyl groups have been delivered to opposite faces of the
molecule in each case (cf. the reaction of 2 with allylmagnesium
bromide).
R
O
RMgX
O
O
R
Diallyl-substituted compounds such as 24d are also poten-
tially available directly from their corresponding dicarbonyl
precursors using the allyltributyltin–titanium tetrachloride
method.7 Thus, the desired products, 24d (51%), 25d (50%) and
26 (25%), were obtained from the corresponding dicarbonyl
compounds 1, 7 and 12c respectively though the yields were
generally lower than above. Moreover, in the case of the
keto aldehyde 12e the reaction stopped at the stage of the
mono-allylation and the alcohol 27 was isolated in 71% yield
(Scheme 6).
O
8
25
Product
RMgX
Yield (%)
cis:trans
25a
25b
25c
25d
MeMgl
74
69
82
73
42.58
50:50
29:71
50:50
EtMgBr
PhMgBr
allyl MgBr
Reaction of the trioxepane with triethylsilane in the presence of
trimethylsilyl triflate
Treatment of the trioxepane 2 with triethylsilane in the presence
of trimethylsilyl triflate (TMSOTf) gave the expected cyclic
ether 28 in good yield (Scheme 7).8 In a similar fashion, the
trioxepanes 8,13c–e were smoothly reduced to the cyclic ethers
29,30c–e respectively (Table 5). However, the cyclic ethers 28–30
were found to be more conveniently prepared by reduction
of dicarbonyl compounds with triethylsilane. The yields were
generally superior to those obtained from the reactions of the
corresponding trioxepanes. Since under similar conditions,
aliphatic and aromatic carbonyl compounds are easily reduced
to give the corresponding alcohols,9 and ω-hydroxy-substituted
ketones give the corresponding cyclic ethers,10 the formation of
cyclic ethers 28–30 from dicarbonyl compounds is most likely
to proceed via the corresponding diols.
Treatment of the trioxepane 2 with a series of Grignard
reagents gave in each case the corresponding dialkyl-substituted
cyclic ethers 24a–d respectively (Table 3). The reaction was
highly stereoselective affording the c-isomer as the predominant
or exclusive product. Product stereochemistry was assigned
1
on the basis of H NMR NOE measurements as for c-24b.
Surprisingly, the analogous reactions involving trioxepane 8
were non-stereoselective; roughly 1:1 mixtures of cis- and
trans-25 were obtained (Table 4). These observed differences in
stereoselectivity between the two trioxepanes, 2 and 8, would
be consistent with the formation of discrete carbocations
from 8 which could subsequently undergo intermolecular
nucleophilic attack on either face in a relatively unselective
fashion.
Experimental
TiCl4-Catalyzed reaction of the trioxepanes with allyltributyltin
Since treatment of acetals with allyltributyltin–titanium tetra-
chloride has been shown to provide allyl-substituted ethers,6 the
analgous reaction with 1,3,5-trioxepanes was examined as an
alternative route to the corresponding diallyl-substituted cyclic
ethers. Thus, the diallyl-substituted cyclic ethers 24d, 25d (83%)
and 26 (73%) were obtained in acceptable yield from the corres-
General
1H and 13C NMR spectra were obtained in CDCl3 with SiMe4
as standard, using a JEOL JNM-EX-270 spectrometer; J values
are given in Hz. IR Spectra were recorded on a Hitachi 260
spectrometer. 8-Benzoylnaphthalene-1-carbaldehyde 1, phen-
anthrene-4,5-dicarbaldehyde 7, 2-(o-benzoylphenyl)acetalde-
J. Chem. Soc., Perkin Trans. 1, 1998
2357

View Article Online
with diethyl ether–benzene (3:97) gave the keto aldehyde 12c
O
Ph
O
(2.76 g, 73%).
2-(o-Benzoylphenyl)-2-methylpropanal 12c. Pale yellow liquid
(Found: C, 81.2; H, 6.2. C17H16O2 requires: C, 81.0; H, 6.3%);
δH 1.30 (6 H, s), 7.2–7.8 (9 H, m) and 9.38 (1 H, s); νmax/cmϪ1
1670 and 1730.
1
o-(1-Benzoylcyclopentyl)benzaldehyde 12e. (85%) Mp 75–
76 ЊC (from hexane) (Found: C, 82.2; H, 6.6. C19H18O2 requires:
C, 82.0; H, 6.5%); δH 1.7–2.0 (m, 4 H), 2.2–2.4 (m, 2 H), 2.6–2.8
(m, 2 H), 7.1–7.8 (m, 9 H), 10.06 (s, 1 H); νmax/cmϪ1 1770, 1680,
1450, 1235, 755 and 705.
Et3SiH
TMSOTf
Method B
O
O
Ph
O
Reaction of dicarbonyl compounds with ethylene glycol in the
presence of chlorosulfonic acid
Ph
O
TMSOTf
Et3SiH
The reaction of the keto aldehyde 1 is representative. A solution
of keto aldehyde 1 (542 mg, 2.08 mmol), ethylene glycol
(10 cm3) and chlorosulfonic acid (281 mg, 2.41 mmol) in di-
chloromethane (40 cm3) was stirred at room temperature for
15 h. After addition of diethyl ether (100 cm3), the organic
layer was washed with aqueous NaHCO3, saturated brine, and
dried over anhydrous MgSO4. After evaporation of the solvent,
the crude product was purified by column chromatography on
silica gel. Elution with diethyl ether–hexane (3:97) gave the
trioxepane derivative 2 (493 mg, 78%).
Method A
28
2
R1
O
O
R4
3,4-Dihydro-1-phenyl-1,6-epoxy-1H,6H-naphtho[1,8-fg][1,4]-
dioxonine 2. Mp. 159–161 ЊC (from ethyl acetate–hexane)
(Found: C, 78.8; H, 5.2. C20H16O3 requires: C, 78.9; H, 5.3%);
δH 3.6–4.4 (4 H, m), 6.50 (1 H, s) and 6.8–8.0 (11 H, m);
δC 66.47, 66.92, 96.01, 97.10, 122.20, 123.02, 124.33, 125.53,
125.95, 127.29, 128.01, 128.16, 128.43, 128.53, 128.61, 130.40,
132.02, 132.18, 132.43 and 135.50.
1,3,4,6-Tetrahydro-1,6-epoxy-phenanthro[4,5-fgh][1,4]-
dioxecine 8. Mp 114–116 ЊC (from methanol) (Found: C, 77.8;
H, 5.0. C18H14O3 requires: C, 77.8; H, 5.1%); δH 3.93 (4 H, s),
6.17 (2 H, s) and 7.2–7.9 (8 H, m); δC 66.84, 102.27, 126.03,
126.67, 127.60, 128.03, 129.83, 134.52 and 137.71.
3,4,6,7-Tetrahydro-7,7-dimethyl-1-phenyl-1,6-epoxy-1H-
benzo[1,4]dioxonine 13c. Colourless oil (Found: C, 77.3; H, 7.0.
C19H20O3 requires: C, 77.0; H, 6.8%); δH 1.30 (3 H, s), 1.40 (3 H,
s), 3.4–3.7 (2 H, m), 3.87 (1 H, d, J 12), 4.29 (1 H, dd, J 12 and
10), 5.0 (1 H, s) and 7.1–7.5 (9 H, m); δC 23.29, 28.25, 38.24,
66.45, 66.76, 100.86, 102.66, 125.12, 125.80, 126.04, 126.27,
127.44, 127.92, 128.10, 128.52, 128.63, 133.21, 143.49 and
144.91.
R3
R2
30
29
R1
Ph
R2
R3
R4
H
H
H
H
a
Me
Ph
Ph
H
H
b
c
Me Me
H
d
e
H
H
Ph
Ph
Ph
-(CH2)4-
Scheme 7
Table 5 Cyclic ethers from the reaction of trioxepanes (Method A)
or dicarbonyl compounds (Method B) with triethylsilane–trimethylsilyl
triflate
3,4,6,7-Tetrahydro-6,7-diphenyl-1,6-epoxy-1H-benzo[1,4]-
dioxonine 13d. Mp 170–172 ЊC (from diethyl ether–hexane)
(Found: C, 79.9; H, 5.8. C23H20O3 requires: C, 80.2; H,
5.9%); δH 3.3–3.8 (4 H, m), 4.13 (1 H, s), 6.47 (1 H, s)
and 6.6–7.5 (14 H, m); δC 55.14, 65.23, 95.76, 102.80, 125.87,
126.93, 127.05, 127.38, 127.50, 127.63, 127.80, 127.92, 128.30,
128.66, 128.97, 129.71, 130.02, 131.73, 138.45, 139.42 and
140.08.
3,4,6,7-Tetrahydro-6-phenyl-1,6-epoxy-1H-benzo[1,4]-
dioxonine-7-spirocyclopentane 13e. Mp 165–167 ЊC (Found: C,
77.8; H, 6.85. C21H22O3 requires: C, 78.2; H, 6.9%); δH 0.7–2.3
(8 H, m), 3.2–3.9 (4 H, m), 6.23 (1 H, s) and 7.2–7.8 (9 H, m);
δC 25.52, 27.38, 33.43, 38.47, 54.10, 64.75, 65.33, 95.42, 105.17,
125.56, 125.64, 127.52, 127.61, 128.09, 128.75, 128.91, 128.99,
129.19, 130.62, 138.83, 146.90.
Phenyl o-(1,3-dioxolan-2-ylmethyl)phenyl ketone 14a. Color-
less oil (Found: C, 76.0; H, 6.1. C17H16O3 requires: C, 76.1;
H, 6.0%); δH 3.14 (2 H, d, J 4.5), 3.6–3.8 (4 H, m), 5.05 (1 H, t,
J 4.5) and 7.2–7.9 (9 H, m); δC 37.27, 64.51, 103.97, 125.73,
127.08, 128.73, 129.79, 129.88, 131.84, 132.79, 134.97, 137.66,
139.17 and 197.96.
Phenyl o-[1-(1,3(dioxolan-2-yl)ethyl]phenyl ketone 14b. Oil
(Found: C, 76.5; H, 6.5. C18H18O3 requires: C, 76.6; H, 6.4%);
δH 1.25 (3 H, d, J 6), 2.8–3.8 (5 H, m), 4.87 (1 H, d, J 4)
and 6.7–8.1 (9 H, m); νmax/cmϪ1 710, 930, 1100, 1460, 1670 and
2950.
Method A
Substrate
Method B
Substrate
Product
Yield (%)
Yield (%)
28
29
30a
30b
30c
30d
30e
2
8
—
—
13c
13d
13e
97
89
—
—
57
53
42
1
7
12a
12b
12c
12d
12e
98
92
71
63
82
59
69
hyde 12a, 2-(o-benzoylphenyl)propanal 12b, and o-(1,2-
diphenyl-2-oxoethyl)benzaldehyde 12d were prepared by the
reported methods.11
Preparation of 2-(o-benzoylphenyl)-2-methylpropanal 12c and
o-(1-benzoylcyclopentyl)benzaldehyde 12e
The preparation of the dialdehyde 12c is representative. Into
a solution of 1,1-dimethyl-3-phenylindene (3.31 g, 15 mmol)
in dichloromethane, was passed a slow stream of ozone
(1.5 equiv.). After evaporation of the solvent, the residue
was dissolved in benzene, and treated with triphenylphosphine
(1 equiv.) for 15 h. After concentration, the crude product
was purified by column chromatography on silica gel. Elution
2358
J. Chem. Soc., Perkin Trans. 1, 1998

View Article Online
Reaction of dicarbonyl compounds with trimethylene glycol in the
presence of chlorosulfonic acid
1.00 (3 H, s), 2.8–4.1 (4 H, m), 5.67 (1 H, s) and 7.3–8.2 (11 H,
m); δC 21.72, 23.11, 30.04, 99.13, 123.19, 125.53, 125.70, 125.86,
127.48, 128.04, 128.46, 128.67, 130.25, 130.95, 132.08, 132.55,
133.99, 135.21, 136.90, 137.84 and 196.12; νmax/cmϪ1 730, 790,
840, 1110, 1290, 1670, 2980.
4,5-Bis(5,5-dimethyl-1,3-dioxan-2-yl)phenanthrene 10. Mp
245–250 ЊC (from methanol) (Found: C, 76.5; H, 7.45. C26H30O4
requires: C, 76.8; H, 7.45%); δH 0.67 (6 H, s), 1.67 (6 H, s),
3.67 (4 H, s), 3.90 (4 H, s), 6.23 (2 H, s) and 7.2–8.2 (8 H,
m); δC 21.71, 23.13, 30.26, 100.66, 125.87, 126.10, 126.54,
127.09, 128.46, 133.67 and 136.96; νmax/cmϪ1 840, 1040 and
1300.
The reaction of the keto aldehyde 1 is representative. A solution
of keto aldehyde 1 (194 mg, 0.75 mmol), trimethylene glycol
(573 mg, 7.5 mmol) and chlorosulfonic acid (144 mg, 1.24
mmol) in dichloromethane (40 cm3) was stirred at room
temperature for 15 h. After the work-up as described above,
the crude products were isolated by column chromatography
on silica gel. The first fraction (elution with diethyl ether–
hexane, 3:97) afforded the 1,3-dioxane derivative 4 (72 mg,
30%). From the second fraction (diethyl ether–hexane, 10:90)
was obtained the trioxepane derivative 3 (80 mg, 34%).
4,5-Dihydro-1-phenyl-1,7-epoxy-1H,3H,7H-naphtho[1,8-gh]-
[1,5]dioxocine 3. Mp 178–179 ЊC (Found: C, 79.1; H, 5.6.
C21H18O3 requires: C, 79.3; H, 5.7%); δH 1.47 (1 H, m), 2.17
(1 H, m), 4.10 (4 H, m), 6.29 (1 H, s), 7.1–7.9 (11 H, m);
δC 30.87, 64.84, 66.60, 97.06, 101.21, 123.58, 123.85, 125.49,
125.87, 125.98, 127.12, 127.80, 128.04, 128.19, 128.29, 128.60,
129.51, 132.54, 133.41 and 142.88.
Phenyl 8-(1,3-dioxan-2-yl)-1-naphthyl ketone 4. Mp 122–
124 ЊC (Found: C, 79.05; H, 5.8. C21H18O3 requires: C, 79.3;
H, 5.7%); δH 1.12 (1 H, m), 1.82 (1 H, m), 3.6–3.8 (4 H, m), 5.62
(1 H, s), 7.3–8.0 (11 H, m); δC 25.39, 65.79, 98.98, 123.13,
125.54, 127.35, 128.00, 128.27, 128.50, 130.19, 130.93, 132.00,
132.47, 134.11, 135.11, 136.87, 137.83 and 195.99; νmax/cmϪ1
1100, 1290, 1660 and 2850.
Reaction of dicarbonyl compounds with methanol in the presence
of chlorosulfonic acid
The reaction of the keto aldehyde 1 is representative. A solution
of keto aldehyde 1 (300 mg, 1.15 mmol), methanol (10 cm3)
and a few drops of conc. HCl in dichloromethane (40 cm3)
was stirred at room temperature for 15 h. After the work-up
as above, the crude products were separated by column
chromatography on silica gel. Elution with benzene–hexane
(1:9) gave the dihydropyran 12a (143 mg, 41%).
1,3-Dimethoxy-1-phenyl-1H,3H-naphtho[1,8-cd]pyran
18.
Mp 152–154 ЊC (Found: C, 78.4; H, 5.9. C20H18O3 requires: C,
78.4; H, 5.9%); δH 3.50 (3 H, s), 3.67 (3 H, s), 6.30 (1 H, s) and
7.2–8.0 (m, 11 H).
1,7-Epoxy-1,4,5,7-tetrahydro-3H-phenanthro[4,5-ghi]-
[1,5]dioxacycloundecine 9. Mp 110–112ЊC (Found: C, 77.6; H,
5.1. C19H16O3 requires: C, 77.8; H, 5.1%); δH 1.9–2.2 (2 H, m),
3.8–4.3 (4 H, m), 5.94 (1 H, s), 6.45 (1 H, s) and 7.5–7.9 (8 H,
m); δC 14.11, 15.28, 32.60, 65.84, 67.30, 67.76, 68.20, 90.24,
99.32, 101.60, 121.02, 124.78, 125.70, 125.91, 127.58, 128.05,
128.28, 128.88, 130.01, 133.78, 134.47, 138.72 and 139.41.
4,5,7,8-Tetrahydro-1-phenyl-8,8-dimethyl-1,7-epoxy-1H,3H-
benzo[1,5]dioxecine 15. Colourless oil (Found: C, 77.9; H,
7.1. C20H22O3 requires C, 77.4; H, 7.1%); δH 1.40 (6 H, s),
1.4–1.6 (1 H, m), 2.0–2.2 (1 H, m), 3.8–4.2 (4 H, m), 4.84
(1 H, s) and 7.1–7.5 (9 H, m); δC 23.38, 30.93, 31.54, 36.73,
62.45, 68.93, 99.52, 105.32, 124.83, 124.93, 126.09, 126.85,
127.80, 127.94, 128.09, 128.16, 128.27, 135.22, 139.32 and
144.08.
Phenyl o-(1,3-dioxan-2-ylmethyl)phenyl ketone 16. Colour-
less oil (Found: C, 76.2; H, 6.55. C18H18O3 requires: C, 76.6;
H, 6.4%); δH 1.1–1.3 (1 H, m), 1.7–1.9 (1 H, m), 2.9 (2 H, d,
J 7), 3.4–3.6 (2 H, m), 3.7–3.9 (2 H, m), 4.57 (1 H, t, J 7) and
7.1–7.8 (9 H, m); δC 25.45, 38.78, 66.52, 101.78, 125.64, 128.09,
128.39, 129.88, 130.25, 132.13, 132.83, 135.29, 137.75, 139.26
and 198.31.
1,3-Dimethoxy-1,3-dihydrophenanthro[4,5-cde]oxepine
19.
Colourless oil (Found: C, 76.8; H, 5.75. C18H16O3 requires:
C, 77.1; H, 5.75%); δH 3.67 (6 H, s), 5.64 (2 H, s) and 7.3–7.9
(8 H, m).
3,4-Dihydro-1,3-dimethoxy-4,4-dimethyl-1-phenyl-1H-
benzo[c]pyran 20. Colourless oil (Found: C, 76.7; H, 7.8.
C19H22O3 requires: C, 76.5; H, 7.4%); δH 1.40 (6 H, s), 3.41 (3 H,
s), 3.62 (3 H, s), 4.87 (1 H, s) and 7.1–7.8 (9 H, m); δC 23.40,
23.60, 38.24, 50.53, 57.50, 101.96, 102.53, 125.48, 126.11,
127.22, 127.67, 127.94, 128.27, 128.37, 135.67, 141.19 and
142.34.
Reaction of 1,3,5-trioxepanes with Grignard reagents
The reaction of keto aldehyde 1 with isopropylmagnesium
bromide is representative. To a solution of isopropylmagnesium
bromide, prepared from isopropyl bromide (1234 mg, 10.04
mmol) and magnesium (244 mg, 10.04 mmol) in diethyl
ether (50 cm3), was added a solution of the trioxepane 2
(304 mg, 1.00 mmol) in benzene (50 cm3) and the resulting
mixture was heated at reflux for 4 h. The reaction mixture was
poured into ice-cold aqueous HCl, and the organic layer
was washed with aqueous NaHCO3, saturated brine and dried
over anhydrous MgSO4. After evaporation of the solvent, the
dihydropyran 24b (94 mg, 28%) was isolated from the crude
product mixture by column chromatography on silica gel
eluting with benzene–hexane (3:7).
1,2-Diphenyl-2-[o-(1,3-dioxan-2-yl)phenyl]ethan-1-one
17.
Mp 162–163 ЊC (Found: C, 80.3; H, 6.2. C24H22O3 requires
C, 80.4; H, 6.2%); δH 1.3 (1 H, d, J 14), 2.0–2.2 (1 H, m), 3.7–4.0
(3 H, m), 4.15 (1 H, dd, J 12 and 5), 5.48 (1 H, s), 6.77 (1 H,
s), 7.0–7.6 (12 H, m) and 8.00 (2 H, d, J 7); δC 25.48, 55.17,
67.23, 67.40, 101.96, 126.86, 126.99, 127.24, 128.39, 128.95,
129.51, 130.58, 132.54, 135.53, 136.95, 137.39, 138.80 and
198.35.
1,3-Dimethyl-1-phenyl-1H,3H-naphtho[1,8-cd]pyran 21.
A
mixture of cis- and trans-isomers 88:12, colourless oil (Found:
C, 87.5; H, 6.7. C20H18O requires C, 87.6; H, 6.6%); δH (major)
1.72 (3 H, d, J 7.5), 1.94 (3 H, s), 5.40 (1 H, q, J 7.5) and 7.2–7.8
(11 H, m); δC 20.92, 25.02, 66.74, 78.13, 119.78, 122.61, 125.27,
125.34, 126.07, 126.38, 127.08, 127.33, 127.91, 128.18, 131.86,
132.00, 132.67, 136.57, 140.63 and 146.34; δH (minor) 1.66 (3 H,
d, J 7.5), 2.00 (3 H, s), 4.68 (1 H, q, J 7.5) and 7.2–7.8 (11 H, m);
δC 19.75, 31.23, 66.96, 79.12, 119.67, 125.00, 125.43, 125.73,
126.67, 126.94, 127.49, 128.09, 128.28, 128.46, 128.66, 131.41,
132.94, 137.02, 137.36 and 144.96.
1,3-Dimethyl-1,3-dihydrophenanthro[4,5-cde]oxepine 25a. A
mixture of cis- and trans-isomers 42:58, mp 94–102 ЊC (Found:
C, 86.8; H, 6.6. C18H16O requires C, 87.1; H, 6.5%); δH (major
isomer) 1.65 (6 H, d, J 7.5), 4.63 (2 H, q, J 7.5) and 7.5–7.9 (8 H,
m); δC 19.28, 71.77, 124.31, 126.34, 126.47, 127.13, 129.43,
132.00, 132.24 and 138.67; δH (minor one) 1.40 (6 H, d, J 7.5),
Reaction of dicarbonyl compounds with 2,2-dimethylpropane-
1,3-diol in the presence of chlorosulfonic acid
The reaction of the dialdehyde 7 is representative. A solution
of dialdehyde 7 (200 mg, 0.87 mmol), 2,2-dimethylpropane-
1,3-diol (1.4 g, 13 mmol) and chlorosulfonic acid (100 mg,
0.87 mmol) in dichloromethane (40 cm3) was stirred at room
temperature for 15 h. After the work-up as above, the crude
products were separated by column chromatography on silica
gel. Elution with benzene gave the keto acetal 10 (214 mg,
61%).
Phenyl 8-(5,5-dimethyl-1,3-dioxan-2-yl)-1-naphthyl ketone 5.
Mp 170–175 ЊC (from ethyl acetate–hexane) (Found: C, 79.6;
H, 6.45. C23H22O3 requires: C, 79.7; H, 6.4%); δH 0.75 (3 H, s),
J. Chem. Soc., Perkin Trans. 1, 1998
2359

View Article Online
5.27 (2 H, q, J 7.5) and 7.5–7.9 (8 H, m); δC 22.61, 76.25, 123.95,
Reaction of 1,3,5-trioxepanes with allyltributyltin in the presence
126.04, 126.97, 127.55, 128.27, 133.59 and 142.57.
of titanium tetrachloride
1,3-Diethyl-1-phenyl-1H,3H-naphtho[1,8-cd]pyran 24a.
A
The reaction of the trioxepane 2 is representative. To a solution
of the trioxepane 2 (255 mg, 0.84 mmol) in dichloromethane
(20 cm3) was added TiCl4 (172 mg, 0.91 mmol), and then
allyltributyltin (805 mg, 2.43 mmol) at Ϫ70 ЊC under nitrogen.
After 1 h, diethyl ether (50 cm3) was added, and the mixture was
washed with aqueous NaHCO3, saturated brine, and dried over
anhydrous MgSO4. After evaporation of the solvent, the crude
products were separated by column chromatography on silica
gel. Elution with benzene–hexane (3:7) gave the dihydropyran
c-24d (134 mg, 49%).
1,c-3-Diallyl-r-1-phenyl-1H,3H-naphtho[1,8-cd]pyran c-24d.
Colourless oil (Found: C, 88.4; H, 7.2. C24H22O requires
C, 88.3; H, 6.8%); δH 2.8–3.2 (4 H, m), 5.0–5.2 (4 H, m), 5.41
(1 H, t, J 6.3), 5.8–6.1 (2 H, m) and 6.8–7.8 (11 H, m);
δC 39.00, 41.13, 69.74, 79.66, 117.05, 117.72, 120.38, 122.66,
125.28, 125.32, 125.55, 126.00, 126.43, 126.54, 127.17, 127.57,
128.00, 133.03, 133.71, 134.39, 134.50, 134.83, 139.44 and
143.86.
1,3-Dihydro-1,3-diallylphenanthro[4,5-cde]oxepine 25d. A 1:1
mixture of cis- and trans-isomers, colourless oil (Found: C,
87.7; H, 6.65. C22H20O requires C, 88.0; H, 6.7%); δH 2.3–2.9
(4 H, m), 4.70 (1 H, t, J 6.3), 4.8–5.0 (4 H, m), 5.16 (1 H, t,
J 6.3), 5.6–5.8 (2 H, m) and 7.5–7.9 (8 H, m); δC 38.35, 40.61,
76.91, 80.15, 116.50, 116.60, 124.58, 125.32, 126.00, 126.38,
126.61, 127.10, 127.31, 127.66, 133.41, 135.20, 135.54, 138.26
and 140.74.
3,4-Dihydro-1,c-3-diallyl-4,4-dimethyl-r-1-phenyl-1H-benzo-
[c]pyran c-26. Colourless oil (Found: C, 86.35; H, 8.2. C23H26O
requires C, 86.75; H, 8.2%); δH 1.33 (3 H, s) 1.34 (3 H, s), 2.3–
2.5 (2 H, m), 3.15 (2 H, d, J 1.7), 3.85 (1 H, dd, J 9.2 and 2.5),
5.0–5.2 (4 H, m), 5.8–6.1 (2 H, m) and 6.8–7.6 (9 H, m);
δC 24.85, 25.32, 34.95, 36.62, 41.67, 76.46, 80.06, 115.72, 117.47,
125.48, 125.77, 126.45, 126.79, 127.39, 127.96, 134.39, 137.34,
140.31, 142.88 and 145.44.
mixture of cis- and trans-isomers 86:14, colourless oil (Found:
C, 87.6; H, 7.4. C22H22O requires: C, 87.4; H, 7.3%); δH (major
isomer) 1.01 (3 H, t, J 7.3), 1.12 (3 H, t, J 7.3), 2.0–2.4
(4 H, m), 5.22 (1 H, dd, J 6.6 and 3.6) and 7.0–7.8 (11 H,
m); δC 8.30, 9.29, 27.37, 29.47, 70.24, 80.20, 119.84, 122.39,
125.27, 125.30, 126.15, 126.24, 126.33, 126.94, 127.33, 128.00,
128.05, 128.28, 133.10, 134.97, 140.61 and 143.86; δH (minor
isomer) 0.96 (3 H, t, J 7.3), 1.13 (3 H, t, J 7.3), 1.9–2.2 (4 H, m),
4.61 (1 H, dd, J 6.6 and 3.6) and 7.0–7.8 (11 H, m); δC 8.68,
14.13, 26.38, 36.80, 70.56, 81.37, 119.50, 123.11, 124.82, 125.48,
126.67, 126.79, 127.46, 127.80, 133.44 and 134.66.
1,3-Diethyl-1,3-dihydrophenanthro[4,5-cde]oxepine 25b.
A
1:1 mixture of cis- and trans-isomers, mp 88–91 ЊC (Found: C,
86.4; H, 7.3. C20H20O requires C, 86.9; H, 7.3%); δH 0.6–0.8
(6 H, m), 1.6–2.1 (4 H, m), 4.40 (1 H, t, J 6.9), 4.88 (1 H, dd,
J 8.4 and 4.2) and 7.4–7.8 (m, 8 H); δC 10.59, 11.18, 26.90,
29.44, 78.51, 82.01, 124.12, 125.01, 125.82, 126.29, 126.51,
126.99, 127.10, 127.40, 129.83, 131.57, 132.51, 133.44, 138.96
and 141.80.
1,c-3-Diisopropyl-r-1-phenyl-1H,3H-naphtho[1,8-cd]pyran
24b. Colourless oil (Found: C, 86.7; H, 8.1. C24H26O requires C,
87.2; H, 7.9%); δH 0.91 (3 H, d, J 7.2), 0.94 (3 H, d, J 6.9), 1.16
(3 H, d, J 6.9), 1.29 (3 H, d, J 6.9), 2.6–2.7 (1 H, m), 2.8–2.9
(1 H, m), 5.37 (1 H, s) and 6.9–7.8 (11 H, m); δC 15.24, 17.84,
19.37, 20.31, 32.42, 36.30, 75.54, 81.67, 119.80, 123.52, 125.14,
125.30, 125.98, 126.20, 126.50, 127.11, 127.62, 127.71, 128.28,
128.34, 132.81, 134.43, 138.74 and 146.63.
1,1,3-Triphenyl-1H,3H-naphtho[1,8-cd]pyran 24c. Colourless
oil (Found: C, 90.4; H, 6.2. C30H22O requires: C, 90.5; H, 6.1%);
δH 5.74 (1 H, s) and 6.7–7.5 (21 H, m); δC 75.62, 85.45, 122.75,
124.91, 125.34, 125.82, 127.80, 128.14, 128.30, 128.50, 128.70,
128.79, 128.91, 129.63, 132.83, 136.35, 136.87, 141.49, 142.97
and 147.04.
1,3-Diphenyl-1,3-dihydrophenanthro[4,5-cde]oxepine 25c. A
mixture of cis- and trans-isomers 29:71, mp 194–196 ЊC
(Found: C, 89.9; H, 5.5. C28H20O requires C, 90.3; H, 5.4%);
δH (major isomer) 6.34 (2 H, s) and 7.5–8.2 (18 H, m); δC 82.75,
126.02, 126.51, 126.76, 127.15, 127.44, 127.66, 127.78, 127.91,
128.00, 128.27, 129.87, 133.48, 141.44 and 142.16; δH (minor
one) 6.82 (2 H, s) and 7.5–8.2 (18 H, m,); δC 79.12, 132.00,
132.11, 138.85 and 140.61.
Reaction of the keto aldehyde 12e with allyltributyltin in the
presence of titanium tetrachloride
To a solution of the keto aldehyde 12e (279 mg, 1.00 mmol)
in dichloromethane (20 cm3) was added TiCl4 (202 mg, 1.06
mmol), and then allyltributyltin (1030 mg, 3.11 mmol) at
Ϫ70 ЊC under nitrogen. After 1 h, diethyl ether (50 cm3) was
added, and the mixture was washed with aqueous NaHCO3,
saturated brine, and dried over anhydrous MgSO4. After
evaporation of the solvent, the crude product was purified by
column chromatography on silica gel. Elution with diethyl
ether–hexane (13:87) gave the keto-alcohol 27 (249 mg,
78%).
Phenyl 1-[o-(1-hydroxybut-3-enyl)phenyl]cyclohexyl ketone
27. Colourless oil (Found: C, 82.0; H, 7.8. C22H24O2 requires:
C, 82.5; H, 7.55%); δH 0.7–1.4 (8 H, m), 2.0–2.6 (2 H, m), 3.18
(1 H, dd, J 6.8 and 0.5), 5.0–5.2 (2 H, m), 5.71 (1 H, br s),
5.9–6.1 (1 H, m) and 7.2–7.8 (9 H, m); δC 12.31, 14.04, 14.29,
28.65, 40.13, 70.91, 78.38, 116.17, 125.61, 128.70, 130.08,
130.35, 130.96, 133.46, 134.09, 136.35, 137.38, 139.08, 141.99
and 199.24.
Reaction of the dihydropyrans, 13c and 20, with methyl-
magnesium iodide
The reaction of 20 is representative. A solution of the dihydro-
pyran 20 (168 mg, 0.54 mmol) and methylmagnesium iodide
[prepared from methyl iodide (775 mg, 5.6 mmol) and mag-
nesium (132 mg, 5.4 mmol)], in diethyl ether–benzene (50 cm3,
1:1) was heated at reflux for 4 h. After work-up as described
above, the acetal 23 (115 mg, 66%) was isolated by column
chromatography on silica gel, eluting with benzene–diethyl
ether (9:1).
3,4-Dihydro-3-(2-hydroxyethyloxy)-1-phenyl-1,4,4-trimethyl-
1H-benzo[c]pyran 22. Reaction of 13c gave 22 as a colourless
oil (Found: C, 77.3; H, 8.0. C20H24O3 requires C, 76.9; H, 7.7%);
δH 1.37 (6 H, s), 1.98 (3 H, s), 2.40 (1 H, br s), 3.4–3.7 (4 H, m),
4.73 (1 H, s) and 6.8–7.5 (9 H, m); δC 23.99, 26.22, 28.39, 31.57,
61.91, 71.57, 79.25, 103.27, 125.59, 126.94, 127.12, 127.26,
127.42, 127.49, 127.76, 127.96, 128.30, 138.92, 141.74 and
146.77; νmax/cmϪ1 3200–3650.
3,4-Dihydro-3-methoxy-1-phenyl-1,4,4-trimethyl-1H-benzo-
[c]pyran 23. Reaction of 20 gave 23 as a colourless oil (Found:
C, 80.6; H, 7.8. C19H22O2 requires C, 80.8; H, 7.8%); δH 1.35
(6 H, s), 2.00 (3 H, s), 3.15 (3 H, s), 4.60 (1 H, s) and 6.8–7.5
(9 H, m); δC 23.83, 26.47, 28.45, 38.21, 56.19, 78.74, 104.15,
125.46, 125.54, 126.92, 127.03, 127.22, 127.30, 127.85, 139.34,
141.94 and 147.13.
Reaction of the trioxepane 2 with triethylsilane in the presence of
TMSOTf
The reaction of the trioxepane 2 is representative. To a solution
of the trioxepane 2 (304 mg, 1.00 mmol) in dichloromethane
(30 cm3) was added triethylsilane (931 mg, 8.01 mmol) and then
TMSOTf (457 mg, 2.06 mmol) at Ϫ70 ЊC and the mixture was
stirred at room temperature for 45 min. After the work-up as
described above, the dihydropyran derivative 28 (239 mg, 97%)
was isolated by column chromatography [silica gel, eluting with
benzene–hexane (1:1)].
1-Phenyl-1H,3H-naphtho[1,8-cd]pyran 28. Mp 96–98 ЊC
(from ethyl acetate–hexane) (Found: C, 87.6; H, 5.75. C18H14O
2360
J. Chem. Soc., Perkin Trans. 1, 1998

View Article Online
requires C, 87.8; H, 5.7%); δH 5.17 (2 H, s), 5.97 (1 H, s), 6.8–7.8
(11 H, m); δC 67.07, 80.21, 120.11, 122.48, 125.42, 125.49,
125.53, 126.39, 126.75, 127.05, 128.25, 128.55, 128.59, 128.71,
132.60, 132.81, 134.97 and 140.42.
1,3-Dihydrophenanthro[4,5-cde]oxepine 29. Mp 55–57 ЊC
(Found: C, 87.6; H, 5.5. C16H12O requires C, 87.35; H, 5.5%);
δH 4.83 (4 H, s) and 7.3–8.0 (8 H, m); δC 71.93, 126.33, 126.43,
126.95, 127.49, 127.85, 131.65, 132.95 and 136.88.
Crystal structure determination of the 1,3,5-trioxepane 2
The crystal of 2 used for X-ray data collection (approx. di-
mensions 0.2 × 0.3 × 0.4 mm) was grown by slow evaporation
from an ethyl acetate–hexane solution and mounted in a sealed
Lindemann capillary tube.
Crystal data. C20H16O3, M = 304.3, colourless prisms,
monoclinic, space group P21/n (alternative setting of No. 14),
a = 8.4715 (8), b = 22.608 (3), c = 8.7042 (9) Å, β = 117.481 (7)Њ,
V = 1478.9 (3) Å3, Z = 4, Dc = 1.367 g cmϪ3, F(000) = 640,
3,4-Dihydro-4,4-dimethyl-1-phenyl-1H-benzo[c]pyran
30c.
Colourless oil (Found: C, 85.5; H, 7.7. C17H18O requires C,
85.7; H, 7.6%); δH 1.28 (3 H, s), 1.44 (3 H, s), 3.68 (1 H, s), 3.72
(1 H, s), 5.72 (1 H, s) and 6.6–7.4 (9 H, m); δC 25.82, 29.42,
33.68, 75.26, 80.72, 122.26, 125.36, 125.88, 126.13, 127.58,
127.71, 127.98, 128.52, 128.77, 135.87, 142.23 and 143.40.
3,4-Dihydro-3,4-diphenyl-1H-benzo[c]pyran 30d. Colourless
oil (Found; C, 87.8; H, 6.4. C21H18O requires C, 88.1; H, 6.3%);
δH 4.00 (1 H, d, J 3), 4.95 (1 H, d, J 3), 5.05 (1 H, d, J 19), 5.10
(1 H, d, J 19) and 6.6–7.2 (14 H, m); δC 50.19, 69.04, 79.93,
124.00, 125.88, 126.60, 126.81, 126.85, 127.21, 127.58, 130.05,
134.14, 137.07, 140.14 and 140.67.
3,4-Dihydro-3-phenyl-1H-benzo[c]pyran-4-spirocyclopentane
30e. Mp 41–45 ЊC (Found: C, 86.1; H, 7.6. C19H20O requires
C, 86.3; H, 7.6%); δH 0.8–2.1 (8 H, m), 4.52 (1 H, s), 4.68 (1 H,
d, J 19), 4.78 (1 H, d, J 19) and 6.8–7.3 (9 H, m); δC 26.90, 31.52,
37.50, 39.14, 48.43, 67.15, 83.51, 123.45, 125.32, 126.07, 126.61,
127.42, 127.71, 127.94, 128.30, 128.68, 133.12, 138.81 and
145.53.
µ(Mo-Kα) = 0.091 mmϪ1
.
Data collection. The intensity data were collected on Siemens
P4 four-circle diffractometer [temperature 293(2) K; θ range:
1.80 to 25.0Њ; Ϫ1 р h р 10, Ϫ1 р k р 26, Ϫ10 р l р 9] using
graphite monochromated Mo-Kα X-radiation (λ 0.710 73 Å)
and ω-scanning. Of the 2578 unique data [R(int) = 0.037]
measured, 1454 had Fo > 4σ(Fo). The data were corrected
for Lorentz and polarisation effects, but not for absorption.
Structure solution and refinement. The approximate positions
of the non-hydrogen atoms were determined by direct methods
[SHELXS-86] (ref. 14). The structure was refined by full-matrix
2
least-squares methods on F2 (SHELXTL/PC15) using all Fo
data and anisotropic temperature factors for all the non-
hydrogen atoms. All the hydrogen atoms were located on dif-
ference Fourier maps and included in the refinement process
at idealised positions with isotropic temperature factors (1.5
times Uiso of the bonded heavy atom). At convergence, the
discrepancy factors R and wR2 were 0.052 and 0.101 respec-
2
tively. The weighting scheme, w = 1/[σ2(Fo ) ϩ (0.0487 P)2]
2
where P = (Fo ϩ2Fc2)/3, was found to give satisfactory analyses
Reaction of dicarbonyl compounds with triethylsilane in the
presence of TMSOTf
of variance. The final difference Fourier map was essentially
featureless (general noise level less than 0.10 e ÅϪ3) with
largest difference peak and hole of 0.18 and Ϫ0.19 e ÅϪ3
respectively.‡
The reaction of the keto aldehyde 12a is representative. To a
solution of the keto aldehyde 12a (221 mg, 0.99 mmol) in
dichloromethane (30 cm3) was added triethylsilane (929 mg,
7.99 mmol) and then TMSOTf (223 mg, 1.00 mmol) at Ϫ70 ЊC
and the mixture was stirred at room temperature for 45 min.
After the work-up as described above, the crude products were
separated by column chromatography on silica gel. Elution
with benzene–hexane (1:1) gave the dihydropyran derivative
30a (147 mg, 71%).
Crystal structure determination of the 1,3,5-trioxocane 3
The general experimental procedures were as described above
for 2.
Crystal data. C21H18O3, M = 318.4, colourless prisms
(approx. dimensions 0.30 × 0.34 × 0.81 mm from diethyl ether–
hexane), monoclinic, space group P21/c (No. 14), a = 8.5324 (9),
b = 11.1349 (11), c = 16.9196 (13) Å, β = 100.699 (7)Њ, V =
1579.5 (3) Å3, Z = 4, Dc = 1.339 g cmϪ3, F(000) = 672, µ(Mo-
Kα) = 0.089 mmϪ1. The 2758 unique data [R(int) = 0.034])
were measured on a Siemens P4 four-circle diffractometer
[temperature 293(2) K; θ range: 2.20 to 25.0Њ; Ϫ1 р h р 10,
Ϫ1 р k р 13, Ϫ20 р l р 20; Mo-Kα X-radiation (λ 0.710 73
Å); ω–scanning). The structure was solved by direct methods
[SHELXS-86 (ref. 14)] and refined by full-matrix least-squares
methods on F2 (SHELXTL/PC15). At convergence, the dis-
3,4-Dihydro-1-phenyl-1H-benzo[c]pyran 30a.12 Mp 88–90 ЊC
(Found: C, 85.3; H, 6.7. C15H14O requires C, 85.7; H, 6.7%);
δH 2.7–2.9 (1 H, m), 3.0–3.2 (1 H, m), 3.8–4.0 (1 H, m), 4.2–4.3
(1 H, m), 5.73 (1 H, s) and 6.7–7.4 (9 H, m); δC 28.79, 63.90,
79.64, 125.89, 126.57, 126.88, 128.07, 128.39, 128.70, 128.86,
133.80, 137.32 and 142.15.
3,4-Dihydro-4-methyl-1-phenyl-1H-benzo[c]pyran 30b.13 Col-
ourless oil (Found: C, 85.4; H, 7.1. C16H16O requires C, 85.7; H,
7.2%); δH 1.35 (3 H, d, J 7.5), 3.01 (1 H, qt, J 7.5 and 7.5), 3.41
(1 H, dd, J 12.0 and 7.5), 3.95 (1 H, dd, J 12.0 and 7.5), 5.57
(1 H, s) and 6.6–7.3 (9 H, m); δC 21.64, 32.69, 69.92, 80.20,
125.88, 126.69, 127.35, 128.00, 128.07, 128.39, 128.66, 128.84,
136.80, 139.35 and 132.14.
2
crepancy factors R and wR2 (w = 1/[σ2(Fo ) ϩ (0.0433 P)2])
where P = (Fo2 ϩ 2Fc2)/3), were 0.052 and 0.101 respectively for
1739 data with [Fo > 4σ(Fo)]. The final difference Fourier map
was essentially featureless (general noise level less than 0.10 e
Å
Å
Ϫ3) with largest difference peak and hole of 0.15 and Ϫ0.15 e
Ϫ3 respectively.‡
Molecular modelling studies on 1,3,5-trioxepane 2 and related
compounds
The molecular modelling studies were performed on a Power
Computing Power Centre Pro computer equipped with 210
MHz 604e RISC processor, 64 Mb of RAM, 1 Mb L2 cache
and 60 MHz system bus.
Initial structures for the Monte Carlo conformational
searches were generated by a custom written Applescript
which allowed the rotation of the torsional angles of selected
molecular segments by up to 35Њ. Molecular Mechanics
minimizations were then performed using the modified MM2
force field implemented in Chem 3D Pro (ver. 3.5.1). Semi-
empirical calculations were performed with MOPAC93 using
PM3 potential functions.
Crystal structure determination of the keto acetal 4
The general experimental procedures were as described above
for 2.
Crystal data. C21H18O3, M = 318.4, colourless prisms
(approx. dimensions 0.42 × 0.46 × 0.63 mm from diethyl ether–
hexane), orthorhombic, space group Pbca (No. 61), a = 8.5902
(14), b = 11.4363 (13), c = 34.025 (5) Å, V = 3342.6 (8) Å3,
‡ Atomic coordinates, thermal parameters and bond lengths and angles
have been deposited at the Cambridge Crystallographic Data Centre
(CCDC). See Instructions for Authors, J. Chem. Soc., Perkin Trans. 1,
available via the RSC Web page (http://www.rsc.org/authors). Any
request to the CCDC for this material should quote the full literature
citation and the reference number 207/229.
Chem3D Pro and MOPAC93 were purchased from
CambridgeSoft as part of ChemOffice.
J. Chem. Soc., Perkin Trans. 1, 1998
2361

View Article Online
Z = 8, Dc = 1.265 g cmϪ3, F(000) = 1344, µ(Mo-Kα) = 0.084
mmϪ1. The 2888 unique data [R(int) = 0.038]) were measured
on Siemens P4 four-circle diffractometer (temperature 293(2)
Perkin Trans. 2, 1992, 5; (d) W. Y. Lee, C. H. Park, H.-J. Kim and S.
Kim, J. Org. Chem., 1994, 59, 878; (e) K. J. McCullough, T.
Sugimoto, S. Tanaka, S. Kusabayashi and M. Nojima, J. Chem.
Soc., Perkin Trans. 1, 1994, 643.
4 P. Brownbridge and T. H. Chan, Tetrahedron Lett., 1979, 4437.
5 W.-L. Cheng, S.-M. Yeh and T. Y. Luh, J. Org. Chem., 1993, 58,
5576.
6 A. Hosomi and H. Sakurai, Tetrahedron Lett., 1978, 2589.
7 (a) Y. Yamamoto and N. Asao, Chem. Rev., 1993, 93, 2207; (b) C.
Hull, S. V. Mortlock and E. J. Thomas, Tetrahedron Lett., 1987,
28, 5343.
8 T. Tsunoda, M. Suzuki and R. Noyori, Tetrahedron Lett., 1979,
4679.
K;
θ range: 1.20 to 25.0Њ; Ϫ1 р h р 10, Ϫ1 р k р 13,
Ϫ1 р l р 40; Mo-Kα X-radiation (λ 0.710 73 Å); ω–scanning).
The structure was solved by direct methods [SHELXS-86
(ref. 14)] and refined by full-matrix least-squares methods on
F2 (SHELXTL/PC15). At convergence, the discrepancy factors
2
R and wR2 (w = 1/[σ2(Fo ) ϩ (0.0504 P)2 ϩ 0.06 P]) where
P = (Fo2 ϩ 2Fc2)/3), were 0.058 and 0.110 respectively for 1215
data with [Fo > 4σ(Fo)]. The final difference Fourier map was
essentially featureless (general noise level less than 0.10 e ÅϪ3
)
9 S. Hatakeyama, H. Mori, K. Kitano, H. Yamada and M.
Nishiyama, Tetrahedron Lett., 1994, 35, 4367.
10 G. A. Olah, M. Arvanaghi and L. Ohannesian, Synthesis, 1986,
772.
with largest difference peak and hole of 0.18 and Ϫ0.17 e ÅϪ3
respectively.‡
11 T. Sugimoto, M. Nojima and S. Kusabayashi, J. Org. Chem., 1990,
55, 4221.
12 T. Fujisaka, M. Nojima and S. Kusabayashi, J. Org. Chem., 1985,
50, 275.
13 M. Miura, A. Ikegami, M. Nojima, S. Kusabayashi, K. J.
McCullough and S. Nagase, J. Am. Chem. Soc., 1983, 105, 2414.
14 SHELXS-86, G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990,
46, 467.
Acknowledgements
We thank the British Council (Tokyo) for the award of travel
grants to K. J. M. and M. N., and for the award of a fellowship
grant to A. M.
15 SHELXTL/PC (Ver. 5.03), G. M. Sheldrick, Siemens Analytical X-
ray Instruments Inc., Madison, WI, USA.
References
1 T. Mukaiyama and M. Murakami, Synthesis, 1987, 1043.
2 G. Mehta, K. S. Rao and N. Krishnamurthy, Tetrahedron, 1989,
45, 2743.
3 The examples of neighbouring participation in dicarbonyl
compounds: (a) G. A. Molander and P. R. Eastwood, J. Org. Chem.,
1996, 61, 1910; (b) H.-J. Wu, S.-H. Tsai and W.-S. Chung, Chem.
Commun., 1996, 375; (c) K. Bowden and F. P. Malik, J. Chem. Soc.,
Paper 8/02928H
Received 20th April 1998
Accepted 2nd June 1998
2362
J. Chem. Soc., Perkin Trans. 1, 1998