Diyne Cyclization in Camphor Derivatives - Experimental and Theoretical Investigation

Article abstract of DOI:10.1515/znb-1996-1121

Derivatives of 3-oxo-camphorsulfonimide (1) with two phenylethynyl groups in the endo positions at the carbons C-2 and C-3 were prepared, and their reactivity towards halogenes and titanium chloride was studied. In every case, the two ethynyl groups led to the annulation of a five-membered ring to the bicyclo[2.2.1] system in an orientation which depends on the bulkiness of the additional substituent in position 3, NMR studies show thai cationic species like 6 and 8 are the first detectable intermediates. They not only contain the fused five-membered ring but also a bond between its exocyclic methylene carbon and an oxygen atom of the sulfonyl group, thus transferring the positive charge mainly to sulfur. Semiempirical calculations (PM3) suggest two intermediates in the formation of such cations.

Full text of DOI:10.1515/znb-1996-1121

Diyne Cyclization in Camphor Derivatives - Experimental and
Theoretical Investigation
Gabriele Wagner, Christian Heiß, Uwe Verfiirth, Rudolf Herrmann*
Institut für Organische Chemie und Biochem ie der Technischen Universität München,
Lichtenbergstr. 4, D
- 85747 Garching
Dedicated to Prof. Ernest Wenkert on the occasion o f his 70th birthday
Z. Naturforsch. 51b, 1655-1662 (1996); received July 24, 1996
Camphor Derivatives, Diyne Cyclization, Five-M em bered Ring Annulation,
Semiempirical Calculation of Intermediates
Derivatives of 3-oxo-cam phorsulfonim ide (1) with two phenylethynyl groups in the endo
positions at the carbons C-2 and C-3 were prepared, and their reactivity towards halogenes
and titanium chloride was studied. In every case, the two ethynyl groups led to the annulation
of a five-m em bered ring to the bicyclo[2 .2 .1 ] system in an orientation which depends on the
bulkiness of the additional substituent in position 3. N M R studies show that cationic species
like
6 and 8 are the first detectable intermediates. They not only contain the fused five-
membered ring but also a bond betw een its exocyclic m ethylene carbon and an oxygen atom
of the sulfonyl group, thus transferring the positive charge mainly to sulfur. Semiempirical
calculations (PM 3) suggest two interm ediates in the formation of such cations.
Introduction
Results and Discussion
Camphor derivatives play an im portant role in
asymmetric synthesis as chiral starting materials,
auxiliaries, and catalysts[l, 2]. Having developed
camphor derivatives (e.g. 3-hydroxycamphorsul-
tam) as bidentate ligands for chiral Lewis acids by
reduction of both double bonds of the oxoimine
(1)[3], which showed reasonable efficiency in en-
antioselective D iels-A lder and ene reactions[4],
we became interested in further modifications of
this system by the introduction of additional sub-
stituents into the positions C-2 and C-3 of cam-
phor. Increase in bulkiness in these positions
should have a pronounced effect on reactivity and
selectivity in catalytic applications. We decided to
do this by the reaction of carbanions with the oxo-
imine (1) but found that most common alkyl- or
aryllithium and Grignard reagents give only mix-
tures of compounds not of synthetic use. We suc-
ceeded, however, in preparing the corresponding
phenylethynyl derivatives (2)-(4) (Fig. 1) and wish
to report here on aspects of their chemistry.
Synthesis and reactivity o f the diynes
The synthesis of the 2,3-bis(phenylethynyl)-
camphorsultam
derivatives
(2)-(4)
proceeds
smoothly by the reaction of phenylethynyllithium
with the oxoimine (1) followed by conventional
m ethylation of the NH and OH groups (Fig. 1).
Surprisingly, the triple bonds in these compounds
are inert towards catalytic reduction (Pd and Pt)
even under hydrogen pressure (50 bar), and do
not react with metal carbonyls such as Co2(CO )8
either. We therefore studied their reactivity in
more detail and found that they readily react with
polarized or polarizable reagents, e.g. halogens.
U nder the reaction conditions (room tem per-
ature), alkynes are less reactive than alkenes
towards halogens, but the products should be of
the same type (simple addition to the multiple
bond)[5].
However, we found that the expected haloal-
kenes did not form at all. On the basis of their
NM R data, the products were identified as tetra-
cyclic systems, having an additional five-memb-
ered ring annulated to the positions C-2 and C-3
of the camphor skeleton. The formation of five-
m em bered rings from suitably constructed diynes
*
Reprint requests to Dr. Rudolf Herrmann.
D
0932-0776/96/1100-1655 $06.00 © 1996 Verlag der Zeitschrift für Naturforschung. All rights reserved.
Unauthenticated
Download Date | 11/18/19 6:18 AM

1656_______________________________________________ G. Wagner et al. ■Diyne Cyclization in Camphor Derivatives
OMe B r
\
Fig. 1. Synthesis of the phenylacetylene derivatives 2-4 and their reactions with Br2, I2, and TiCl4. Typical NMR
numbering of the carbon atoms is shown in formulae 2, 5, and 9.
via cations seems to be a general reaction, if the
alkynes can arrange such that the distance be- zations occur, they generally involve other carbon
tween carbon atoms which are expected to form
cannot lead to a direct cyclization reaction; if cycli-
atoms in addition to the alkynes, which was ob-
the new bond does not exceed ~ 4 A, with the tor- served in the case of phenylethynyl-substituted
sion angle between the two alkyne groups not ex- semibullvalenes[7].
ceeding ~ 40° (see also the semiempirical calcula-
M echanism o f the cyclization process
tions below). Thus, l,2-bis(phenylethynyl)benzene
(distance 3.65 A , torsion angle 0°) cyclizes in a
completely analogous way[6]. A larger distance
In order to get some insight in the process lead-
ing to different cyclization products, we followed
Unauthenticated
Download Date | 11/18/19 6:18 AM

1657
G. Wagner et al. ■Diyne Cyclization in Camphor Derivatives
(Fig.2), and the signals of the starting material dis-
appear only after more than two hours. An inter-
mediate with a maximum intensity after 20 min-
utes was observed, which showed a typical 8-H
pattern (doublets at d = 4.00 and 4.94 ppm, J =
15.1 Hz, spectrum 4 from bottom ), but with a very
large chemical shift difference, com pared to the
starting material (degenerate signal at 3.34 ppm;
the reaction of the compounds 3 with iodine and 4
with bromine by proton and carbon NMR. Proton
NMR is particularly well suited for this purpose,
as starting materials, interm ediates and products
have well separated and easy to identify signals
between d = 2.5 and 5.0 ppm. Typical resonances
in this region involve OH (broad singlet), N-
methyl and O-methyl (singlets), and the diastereo-
topic protons 8-H (two doublets with J = 11-15 Hz; Fig. 2, bottom). This points to a severe structural
for NM R numbering of the atoms, see Fig. 1). In
the case of the starting materials 2 and 3, the pro-
tons 8-H are degenerate and appear as singlets.
For concentrations of 0.2 mol/1 in CDC13, the ki-
netics can be conveniently monitored. W herever
interm ediates with sufficient lifetime occurred, we
confirmed our assignments by 13C NM R and two-
dimensional techniqes (COSY, HM QC, HMBC).
'H -N M R spectra are collected in Table I and 13C-
NM R spectra in Table II.
change in the environment of the S 0 2 group which
strongly influences the electronic properties and
the geometry of the five-membered heterocyclic
ring. The 13C spectrum clearly shows that five-
membered-ring annulation by ring closure of the
alkynyl groups has already occurred. Initial attack
of I+ at C -ll must be responsible for this change,
but no carbon-centered cation can be detected. As
structure for this first intermediate, we therefore
suggest the cation 6, with the positive charge local-
ized mainly on sulfur, and a bond between one
oxygen atom of the S 0 2 group and the exocyclic
carbon atom at the newly formed ring. This struc-
Typical results of our NM R experiments are
shown in Fig. 2 and 3. Iodine reacts comparatively
slowly with the
A^-methylated com pound (3)
Table I. 'H NMR data (d values; multiplicity, integration and coupling constants (Hz) in brackets) of new com-
pounds at 360.134 MHz in CDC13. Atom numbering follows that of the corresponding carbon atoms (Fig. 1).
Phenyl
Other
Comp.
8-H (2 d)
9,10-H (2 s)
1.04, 1.52
4-H (d)
5,6-H (m)
7.25 (m, 6H)
7.38 (m, 4H)
3.15 (s, 1H)
5.20 (s, 1H)
2
2.23 (5.2)
1.85 (1H), 1.94
(1H), 2.11 (1H)
2.30 (1H)
3.38, 3.42 (13.7)
2.80 (s, 1H)
2.93 (s, 3H)
3.34la] (2H)
1.04, 1.62
1.01, 1.59
1.09, 1.67
7.27 (m, 6H)
7.39 (m, 4H)
1.76 (1H), 1.92
(1H), 2.11 (1H)
2.32 (1H)
1.76 (1H), 1.89
(1H), 2.03 (1H)
2.32 (1H)
3
4
5
2.17 (5.1)
2.38 (5.2)
3.04 (4.6)
2.90 (s, 3H)
3.46 (s, 3H)
3.29laI (2H)
7.24 (m. 6H)
7.40 (m, 4H)
6.60 (d. 1H, 7.3)
6.90 (m, 6H)
2.95 (s, 3H)
3.69 (s, 3H)
1.65 (2H), 1.80
(1H), 2.05 (1H)
3.36, 3.46 (13.5)
7.28 (m, 1H)
7.40 (d, 2H, 7.4)
3.33 (s, 3H)
4.10 (s, 1H)
3.27 (s, 3H)
3.84 (s, 3H)
2.61
(s, br, 1H)
2.90 (s, 3H)
2.42 (4.4)
2.87 (4.4)
2.03 (4.8)
1.48 (1H), 2.10
(2H), 2.33 (1H)
1.50 (1H), 2.05
(2H), 2.40 (1H)
1.48 (1H), 1.90
(2H), 1.99 (1H)
4.00, 4.94 (15.1)
4.08. 5.78 (15.4)
2.66, 2.76 (13.5)
1.19, 1.58
6.90-7.35 (m)
6
8
9
1.54
1.60
6.70-7.30 (m)
1.21,
1.03,
7.24 (m, 6H)
7.37 (m, 4H)
10M
6.87 (m)
3.66 (s, 1H)
5.21 (s, 1H)
3.26, 3.35 (13.9)
2.81, 2.95 (13.6)
1.03, 1.60
1.04, 1.56
1.48 (1H), 1.54
(1H), 1.79 (1H)
1.94 (1H)
1.30 (1H), 1.96
(2H), 2.61 (1H)
2.73 (5.2)
2.47 (4.6)
2.87 (s, 3H)
3.68 (s, 3H)
6.80-7.45 (m)
11
H Degenerate signals, appearing as singlet;
13-H: 6.27 (s, 1H).
Unauthenticated
Download Date | 11/18/19 6:18 AM

1658
G. Wagner et al. ■Diyne Cyclization in Camphor Derivatives
Table II. 13C NMR data (d values) of new compounds at 90.556 MHz in CDC13. Numbering of the carbon atoms
follows Fig. 1.
Comp. 1
4
2
3
5. 6
7
8
9. 10
11
12
13
14
Ph (Cq) Ph (CH)
Others
-
2
3
4
62.3 73.0 82.3 56.7 24.2
28.6
49.2 51.5 23.7
24.1
122.0
122.2
128.4
128.5
128.8 128.9
132.0 132.1
56.6 76.5 83.3 57.7 23.9
28.6
49.1
49.2
23.3
23.4
122.1
122.4
128.1 128.2 29.1
128.5 128.6
131.7 131.8
56.2 76.1 88.9 50.1
23.3 49.0 48.6 21.7
28.6
122.1
122.2
127.9
28.7
128.1
23.1
128.3 128.3 52.4
131.5 131.6
89.1
5
8
55.5
97.3 54.8 23.4
27.9
51.6 49.3 22.8
24.5
134.2
138.2
126.7 127.4 28.9
127.6 128.3 55.4
129.4 129.9
54.9 80.4 97.7
23.4 52.4
27.4
46.6
49.2 20.9
22.1
133.7
135.6
127.7 128.2 29.9
129.8
129.9 55.8
130.0 130.9
126.7
9
57.6 84.7 94.7 52.5 24.5 52.2 47.9 24.0
138.2
143.9
28.1
126.8
128.4 129.0
129.1
28.6
26.1
131.5
-
79.7
10
11
59.8
90.5 53.8 23.8 51.8 51.1
28.5
22.8
23.2
134.7
137.2
126.8 127.1
127.2 127.6
128.4 129.5
57.0 86.5 98.2 47.1
Not assigned individually.
24.8 51.8 47.0 23.6
28.3
136.2
137.2
126.9
127.3 28.1
128.1 129.1 54.8
130.6
24.7
132.3
ture is supported by the calculations reported be-
low. The cation then reacts with the remaining I"
to the diododialkene (9), which was not only ob-
served in the kinetic experiment (Fig. 2, top; 8-H
doublets at <5 = 2.66 and 2.76 ppm, J = 13.5 Hz),
but could be isolated and characterized. Although
the formation of E /Z isomers with respect to the
exocyclic double bond could be a priori expected,
only one product was observed. The absence of
N O E effects of the phenyl group at this double
bond on protons of the camphor part of the mole-
bromodialkene 11, in close analogy to the reaction
of 3 with iodine. However, there is an additional
product (about 1/3 of the total material) which
formed directly from 4 and not from the cation 8;
NM R spectroscopy showed that it is the cyclized
compound 5 where the initial attack of Br+ oc-
curred at C-13 (Tables I and II). This means that
after cyclization the cation with the charge at C-
12 has no possibility of stabilizing by charge
transfer to sulfur, and hence there is no chance to
observe this quite reactive ion by NM R spectros-
cule suggests that (9) is the Z isomer shown in copy. As soon as the starting material was con-
Fig. 1, i.e. the replacement of oxygen by iodide at
sumed, the amount of 5 in the mixture remained
the double bond occurs with retention of configu- constant, indicating that the pathways leading to
ration. This stereochemical result is commonly ob- either 5 or 11 are not interconnected. We believe
served in vinylic nucleophilic substitution[8, 9],
that the effect of O-m ethylation on the site of the
The reaction of the dimethylated compound 4
attack of the electrophile (C -ll vs. C-13) is mainly
with iodine is extremely slow, and kinetic experi- of steric origin.
ments were therefore performed with bromine
As the products 2-4 are optically pure and of
(Fig. 3). Here the starting material 4 rapidly disap- potential interest as chiral ligands, we have studied
peared. while the typical pattern of the sulfur-sta-
bilized cation 8 appeared and over a period of
some hours converted to the final product, the di-
the complexation behaviour of 2 towards TiCl4.
We expected the form ation of a chelating complex
such as 7. Due to solubility problems it was not
Unauthenticated
Download Date | 11/18/19 6:18 AM

G. Wagner et al. • Diyne Cyclization in Camphor Derivatives
1659
rationale for its formation is that in the chelate
complex 7 the attack of an electrophile at C -ll is
hindered by the metal site, in analogy to the effect
of O-methylation in 4. As HC1 is liberated during
the formation of 7, it is not surprising that it adds
to the triple bonds, whereby the proton attacks C-
13, leading to a reactive cation which is then
trapped at C-12 by the chloride ion.
Sem iem pirical calculation o f cationic intermediates
The conclusions drawn from NM R spectra con-
cerning the structures of the intermediate cations
of the cyclization reactions might seem specula-
tive, and we therefore carried out supporting semi-
empirical calculations (PM3) on the intermediates.
We first checked a simple model, 1,2-diethynyl-
benzene, as starting material. To a geometry-opti-
mized molecule was added
a proton as the
electrophile in a suitable distance to one of the
triple bonds (1.0 to 1.4 A). Geom etry optimization
of the ensemble led to the typical cation with the
indene skeleton, as postulated in the halogenation
5^0
4.5
4.3
3.5
3.0
(ppm)
Fig. 2. Reaction of 3 with iodine. 'H NMR spectra (from
bottom to top) after t = 0, 5, 10, 20, 30, 56 min. 3: d = 3.34
(2H); 6: d = 4.94 (d), 4.00 (d); 9: Ö = 2.76 (d), 2.66 (d).
reactions
of
l,2-bis(phenylethynyl)benzene[6]
with no obvious activation barrier or further inter-
mediates as local energy minima. The optimization
converged more rapidly when the proton was
placed in the plane of the alkyne groups, probably
because of the better overlap of the empty orbital
at the (formally cationic) terminal carbon atom
with the filled jr-orbital of the non-protonated al-
kyne group. We then investigated the conditions
which a starting material has to fulfill in order to
cyclize analogously. We did this by constructing
both cis- and frans-1,2-diethynylcycloalkanes (4- to
7-membered rings). The distances between the
carbon atoms which should form bonds to give the
five-membered annulated ring varied between
3.47 A (c/s-l,2-diethynylcyclobutane) and 4.06 A
time (min)
(frans-1,2-diethynylcycloheptane).
The
corre-
Fig. 3. Reaction of 4 with bromine. Composition of the
reaction mixture according to 'H NMR spectra.
sponding torsion angles varied between 1.2° (al-
most in plane) and 83.0° (for frans-1,2-diethynyl-
cyclobutane). After the addition of a proton in a
possible to follow the reaction by NMR spectros- similar position as described before, the geometry
optimization was performed. Cyclization was ob-
served for all compounds where the distance be-
tween the carbon atoms to be connected did not
exceed the limit of 3.8 A, whereas for trans-1,2-
diethynylcyclobutane (4.71 A), trans-1,2-diethy-
nylcyclopentane (4.05 A), and frans-l,2-diethynyl-
copy, but after aqueous workup of the reaction
mixture, an appreciable amount (~ 60%) of a new
product was obtained, which we identified as the
cyclized chlorinated compound 10, together with
unchanged ligand 2. The compound is analogous
to the byproduct 5 of the brom ination of 4. The
Unauthenticated
Download Date | 11/18/19 6:18 AM

1660
G. Wagner et al. • Diyne Cyclization in Camphor Derivatives
Fig. 4. Calculated intermediate (left), transition state (middle), and final product (right) of the protonation reaction
of l,2-bis(phenylethynyl)benzene. The relative heats of formation are 0, +5.2, and -23.6 kcal/mol. The planes through
the phenyl groups are almost perpendicular to the plane of the disubstituted benzene in all cases.
cycloheptane (4.06 A), no cyclization was ob- cation. The distance between the carbons to be
served. The compounds which did not cyclize
showed higher torsion angles as well (83.0, 41.2,
and 45.0°, respectively). We therefore propose as
a rule-of-thumb that formation of a five-memb-
ered ring from a 1,2-dialkyne should occur in all
cases where a conformation can be achieved which
shows a distance between the carbon atoms to be
connected below ~ 4 A, together with a torsion
angle between the alkyne groups below ~ 40°. The
connected is shortened considerably (2.20 vs.
2.92 A.) The low barrier is in accord with the con-
clusions drawn earlier from the experimental data
concerning the high degree of concertedness of
this cyclization reaction[6].
Next, we optimized the geometry of the cam-
phor derivative 2 and found the distances between
the carbon atoms which could potentially be con-
nected to be 3.23 A (C -ll - C-14) and 3.27 A (C-
12 - C-13) (we use the NM R numbering for the
carbon atoms shown in Fig. 1). Together with a
torsion angle of 9.3° such camphor derivatives are
almost ideal substrates for the cyclization. How-
ever, the direction of the cyclization (i.e the ques-
tion if the prim ary attack of the electrophile occurs
at C -ll or C-13) cannot be predicted from these
geometric considerations, as both distances are al-
most the same. Addition of a proton to C -ll fol-
lowed by geom etry optimization led to the first
interm ediate 12 (Fig. 5) as a local minimum, in
close analogy to the interm ediate in the proton-
ation of l,2-bis(phenylethynyl)benzene. The dis-
tance between the carbon atoms to be connected
(C-12 and C-13) is shortened compared with the
experimentally studied
l,2-bis(phenylethynyl)-
benzene[6] (distance 3.65 A, torsion angle 0°)
obeys this rule-of-thumb. However, it did not cy-
clize during the geometry optimization, but gave a
local minimum with a distance of 2.92 A between
the carbon atoms to be connected (Fig. 4, left), al-
though we calculated the cyclized cation (Fig. 4,
right) to be ~ 24 kcal/mol more stable than this
intermediate. We therefore estimated the activa-
tion barrier for the transition between the two cat-
ions using M o p a c 's transition state ability (saddle
point calculation (keyword s a d d l e ) followed by
optimization of the transition state (keyword t s ) ) .
We found that the transition state is approximately
5 kcal/mol higher in energy than the interm ediate
12
13
14
Fig. 5. Calculated intermediates in the protonation of 2 (PM3). Heats of formation relative to 12 (0 kcal/mol): 13:
-38.2; 14: -64.2 kcal/mol.
Unauthenticated
Download Date | 11/18/19 6:18 AM

1661
G. Wagner et al. • Diyne Cyclization in Camphor Derivatives
was separated and the aqueous layer extracted
with dichloromethane (3x50 ml). TTie combined
organic layers were dried (Na2S 0 4), the solvents
were evaporated, and the residue was recrystal-
lized from dichloromethane/hexane (1:1). Yield
3.9 g (90%), m.p. 224-225 0 C; [«]t>2 = -30.0 (c =
1.0 in acetone); [a] ^ = -34.2 (c = 1.0 in ethanol);
m/z'. 431 [M+],
starting material (2.85 vs. 3.23 Ä). The next step in
the reaction should be the annulation of the five-
membered ring, leading to a second interm ediate
13. This cation is indeed another local energy mini-
mum (38.2 kcal/mol more stable). The distance C-
12 - C-13 is typical for a single bond (1.48 A). In
the final cation 14 the positive charge is stabilized
considerably by transfer to the sulfur atom. This is
reflected in an additional decrease of the heat of
formation (64.2 kcal/mol more stable than 12, 26.0
kcal/mol more stable than 13). The stabilization
effect is also reflected in the decrease of the calcu-
lated dipole moments (12: 8.0 D, 13: 8.0 D, 14: 5.1
D). The S-O distance was calculated as 1.69 A
(typical S-O single bond, com pared with the
double bond of 1.42 A to the second oxygen), and
the newly formed C-O bond as 1.41 A(single
bond). Attem pts to localize the transition states
failed, probably due to the high complexity of the
molecules. Energy minima lower than 14 were not
detected. Thus, these calculations support strongly
our conclusions drawn from N M R measurem ents
concerning the structures of the intermediates.
C26H „N O ,S (431.56)
Calcd
C 72.4 H 5.8 N 3.1% ,
Found C 71.9 H 5.9 N 3.1% .
(]aS,3aS, 7R)-7-H ydroxy-l ,8,8-trim ethyl-l a, 7-
bis(phenylethynyl) 1,1a,4,5,6,7-hexahydro-3H-
3a,6-m ethano-2,1-benzisothiazole 2,2-D ioxide (3):
To a solution of 2 (0.30 g, 0.70 mmol) in 3 ml of
dry THF, sodium hydride (60% in mineral oil, 28
mg, 0.70 mmol) was added. A fter stirring for 30
min at room tem perature, dimethyl sulfate (67 /ul,
0.70 mmol) was added, and stirring was continued
for 12 h. A fter workup with dichloromethane/
water, the residue was recrystallized from dichlo-
romethane/hexane (1:1). Yield: 0.30 g (97%), m.p.
166-168° C; [cc]d = -16.9 (c = 0.6 in ethanol); m /
z: 445 [M+],
C27H ,7NO,S (445.59)
Calcd
C 72.8 H6.1 N 3.1% ,
Experimental
Found C 72.5 H 6.0 N 3.2%.
NMR spectra were obtained with a Bruker AM
360 instrument ('H : 360.13 MHz, 13C: 90.56 MHz).
Optical rotations were m easured with a Perkin El-
mer 241 M polarimeter and mass spectra with a
Varian CH5 instrument (El mode, 70 eV).
Semiempirical calculations were perform ed with
P M 3 using the standard param etrization as im ple-
mented in the H y p e r c h e m (release 3 ) and M o p a c
(version 7.0 for DOS) software packages. H y p -
(1aS,3aS, 7R)-7-M ethoxy-1,8,8-trim ethyl-l a, 7-
bis(phenylethynyl)-! ,1a,4,5,6,7-hexahydro-3H -
3a,6-m ethano-2,l-benzisothiazole 2,2-D ioxide (4):
The compound was prepared as described for 3,
except that the amounts of sodium hydride and
dimethyl sulfate were doubled. The crude product
was purified by chromatography (silicagel/dichlo-
rom ethane). R f 0.32, yield: 0.31 g (98%), m.p. 6 2 -
65° C; [a]o = +16.4 (c = 0.9 in ethanol); m/z:
459 [M +\
e r
c
h
e
m
was used for geometry optimizations (Po-
lak-Ribiere algorithm) followed by single-point
calculations and M o p a c for the localization of
transition states.
C^8H 29NO,S (459.61)
Calcd
C 73.2 H 6.4 N 3.1% ,
Found C 73.5 H 6.2 N 3.0% .
Starting materials 2-4
Reaction o f 2 with TiCl4:
(1aS,3aS,7R)-7-H ydroxy-8,8-dim ethyl-la, 7-
(3aS,6aS,9aS)-7-(Z-2-Chloro-2-
phenylm ethylene)-6a-hydroxy-10,10-dimethyl-8-
phenyl-3a,4,5,6,6a, 7-hexahydro-l H,3H-3a,6-
bis(phenylethynyl)-! ,1a,4,5,6,7-hexahydro-3H-
3a,6-m ethano-2,l-benzisothiazole 2,2-D ioxide (2):
To a solution of phenylacetylene (2.6 g, 25 mmol)
in dry diethyl ether (50 ml), 25 mmol of butyl lith-
ium (2.5 M solution in hexane, 10 ml) was added
dropwise at room tem perature. A fter stirring for
30 min the oxoimine 1 (2.3 g, 10 mmol) was added
in five portions over a period of 10 min. The mix-
ture was heated at reflux for 12 h and then hy-
drolyzed with 25 ml of water. The organic layer
methano-indeno[3a,4-c]isothiazole
2,2-D ioxide
(10): To a solution of TiCl4 (55 /d, 0.50 mmol) in
3 ml of dichloromethane, a solution of the sulfon-
amide 2 (0.22 g, 0.50 mmol) in 12 ml of dichloro-
m ethane was added under nitrogen at 0 °C. The
mixture was stirred for 5 h. W ater (10 ml) was
added, and the aqueous layer was extracted with
Unauthenticated
Download Date | 11/18/19 6:18 AM

1662
G. Wagner et al. ■Diyne Cyclization in Camphor Derivatives
dichlorom ethane (2x10 ml). The combined or-
ganic layers were dried (M gS04), and the solvent
was evaporated. The crude product was purified
by chrom atography (silicagel, dichloromethane)
followed by recrystallization from chloroform. R f
phenyl-3,3a,4,5,6,6a-hexahydro-l H,9H-3a,6-
m ethano-indeno/3a,4-c/isothiazole 2,2-D ioxide (9):
The product was purified by recrystallization from
dichlorom ethane/hexane 1:1. Yield: 0.68 g (97%),
m.p. 95 °C (dec.); [a]o = +25.4 (c = 1.1 in ethanol);
no mass spectrum obtained due to decomposition.
0.13, yield 0.14
g
(60%), m.p. 156-160°
C; [a]f3 = -21.9 (c = 0.8 in chloroform); m/z: 468
[M +l+] (rel. 35C1, FAB spectrum).
C27H 27I^NC>3S (699.39)
Calcd
C 46.4 H 3.9 N 2.0% ,
Found C 46.2 H 3.7 N 2.0% .
C26H ,6ClNO,S (468.02)
Calcd
C 66.7 H 5.6 N 3.0%,
(3aS,6aS,9aS)-9-Brom o-7-(Z-2-brom o-2-
phenylm ethylene)-6a-m ethoxy-l ,10,10-trimethyl-8-
phenyl-3a,4,5,6,6a,7-hexahydro-l H,3H-3a,6-
m ethano-indeno[3a,4-c]isothiazole 2,2-D ioxide (5)
and (3aS, 6aS, 9aR)-7-brom o-9- (Z-2-brom o-2-phe-
nylm ethylene) -6a-m ethoxy-l ,10,10-trimethyl-8-
phenyl-3,3a,4,5,6,6a-hexahydro-l H,9H-3a,6-
Found C 66.5 H 5.6 N 3.1% .
Kinetic N M R measurements
The sulfonamide 3 or 4 (0.10 mmol) was dis-
solved in CDC13 (0.3 ml), and the solution
transfered to an NMR tube, and a solution of io-
dine or bromine (0.10 mmol) in CDCI3 (0.2 ml)
was added to it. NMR m easurements were started
immediately afterwards and repeated first every 5
min and later at larger intervals, until no further
changes of the spectrum occurred during a period
of 1 h.
m ethano-indeno [3a,4-cJisothiazole
2,2-D ioxide
(11): The products were separated by chrom ato-
graphy (silicagel, hexane/ethyl acetate 4:1). 5: Rf.
0.66, yield 0.16 g (24%), m.p. 250-254 0 C (dec.);
[a]o = -163 (c = 0.1 in ethanol/dichlorom ethane
1:1); m /z: 617 [M+l (rel. 79Br).
C28H 79B r2N 03S (619.43)
Calcd
C 54.3 H 4.7 N 2.3% ,
Com pounds 5, 9, and 11
Found C 54.0 H 4.6 N 2.2% .
For preparative variants of the NMR experi-
ments, the tenfold amounts of starting materials
were used, and the solvent CDC13 was replaced by
dichlorom ethane, maintaining the same concen-
trations as in the experiments. A t the end of the
reaction, a 10% aqueous solution of sodium thio-
sulfate (10 ml) was added, and the organic layer
washed with water (2x10 ml). A fter drying of the
organic layer (M gS04) and evaporation of the
solvent, the crude product was purified as de-
scribed below.
11: Rf. 0.55, yield 0.25 g (40%), m.p. 200-
201 °C; [a]o = -12.8 (c = 0.1 in ethanol); m/z: 617
[M+] (rel. 79Br).
C28H 29B r2NC>3S (619.43)
Calcd
C 54.3 H 4.7 N 2.3% ,
Found C 54.1 H 4.6 N 2.3% .
A cknowledgem ents
The authors wish to thank Prof. Dr. Ivar Ugi,
TU M ünchen, for supporting this work. Financial
support by Deutsche Forschungsgemeinschaft is
gratefully acknowledged.
(3aS,6aS,9aR)-6a-Hydroxy-7-iodo-9-(Z-2-
iodo-2-phenylm ethylene)-l,10,10-trim ethyl-8-
[1] W. Oppolzer, Tetrahedron 43, 1969 (1987).
[2] W. Oppolzer. Pure Appl.Chem. 62. 1241 (1990).
[3] A. Ketter, G. Glahsl, R. Herrmann, J.Chem.Res. (S)
278 (1990); J.Chem.Res. (M) 2118 (1990).
[4] A. Ketter. R. Herrmann. Z.Naturforsch. 45b, 1684
(1990).
[5] G. Modena, F. Rivetti. U. Tonellato, J.Org.Chem. 43,
1521 (1978).
[6] H. W. Whitlock (Jr.), P E. Sandvick. J.Am.Chem.
Soc. 88,4525 (1966).
[7] H. Quast. J. Carlsen. H. Röschert, E.-M. Peters. K.
Peters, H. G. von Schnering, Chem.Ber. 125. 2591
(1992).
[8] Z. Rappoport. Adv.Phys.Org.Chem. 7. 1 (1969).
[9] S. I. Miller. Tetrahedron 33. 1211 (1977).
[10] G. Glahsl, R. Herrmann, J.Chem.Soc., Perkin Trans.
1 1753 (1988).
[11] F. A. Davis, A. Kumar, B.-C. Chen. J.Org.Chem. 56.
1143 (1991).
Unauthenticated
Download Date | 11/18/19 6:18 AM