The first silatropylium ion stabilized by rigid σ-frameworks: Preparation, properties, and some reactions

Article abstract of DOI:10.1016/S0040-4020(01)00253-8

The first cyclic π-conjugated silylium ion, the silatropylium ion 5 annelated with three bicyclo[2.2.2]octene units having a mesityl group on the silicon atom, was prepared with tetrakis(pentafluorophenyl)borate (TPFPB) as a counter anion in CD₂

Full text of DOI:10.1016/S0040-4020(01)00253-8

TETRAHEDRON
Pergamon
Tetrahedron 57 *2001) 3645±3656
The ®rst silatropylium ion stabilized by rigid s-frameworks:
preparation, properties, and some reactions
Tohru Nishinaga, Yoshiteru Izukawa and Koichi Komatsup
Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan
Received 2 October 2000; accepted 21November 2000
AbstractÐThe ®rst cyclic p-conjugated silylium ion, the silatropylium ion 5 annelated with three bicyclo[2.2.2]octene units having a
mesityl group on the silicon atom, was prepared with tetrakis*penta¯uorophenyl)borate *TPFPB) as a counter anion in CD2Cl2 at 2508C by
hydride abstraction from the corresponding silacycloheptatriene *silepin). The observed chemical shifts of 1H, 13C, and 29Si NMR for cation 5
were in agreement with the calculated values. These results and NICS calculations indicated that cation 5 has the aromaticity approaching
that of the tropylium ion. When cation 4 having a methyl group instead of a mesityl group and cation 5 were generated in the presence of two
equivalents of acetonitrile, the acetonitrile complexes 16 and 17 with considerable covalent-bond character were formed. For the complexes
16 and 17, the exchange process of acetonitrile was observed for the ®rst time as an acetonitrile complex of a silyl cation, and the
intermediacy of a pentavalent silicon was demonstrated. When cations 4 and 5 were generated by the use of triphenylmethyl perchlorate
instead of triphenylmethyl TPFPB, covalent perchlorate esters 18 and 19 having the O±Si covalent bond were formed instead of ionic salts.
By treatment with water, the perchlorate ester 18 was found to undergo a novel rearrangement to give a cyclopentene derivative 20a spiro-
connected with bicyclo[2.2.2]octane at the 3-position. q 2001Elsevier Science Ltd. All rights reserved.
1. Introduction
Then a question arises if the silylium ion can be stabilized
by the aromatic conjugation with 2p orbitals of carbons. The
recent achievement of silaaromatic compounds15 such as
silabenzene,16 silanaphthalene,17 and silacyclopentadienyl
dianion18,19 has revealed that the aromatic conjugation
between the 3p orbital of silicon and 2p orbitals of carbons
is possible. However, there has been no study to con-
vincingly answer the question as to the cationic conjugated
systems involving the silylium ion. The early theoretical
calculations at a very low level on silacyclopropenylium
ion indicated the absence of any signi®cant aromatic stabili-
zation.20 However, in the same paper, the silacyclopenta-
dienyl anion was also shown to be non-aromatic,20 which
contradicts the recent results of calculations at a higher
level.21 More adequate calculation method would be
required for a reliable conclusion on the aromaticity of sila-
cyclopropenylium ion to be reached.
From the viewpoint of the comparison between the chem-
istry of carbon and that of silicon, a trivalent silyl cation *a
silylium ion) has attracted great interest of the organic and
organometallic chemists, and many efforts have been
devoted to synthesize the silylium ion in condensed phases
during the last decades.1 However, most of the claims on
the proof for the silylium ion2±4 were challenged.5±11 For
example, much debate8±11 has been made whether [Et3Si
2
*toluene)]1B*C6F5)4
prepared by Lambert3 and i-
Pr3Si1CB11H6Br6 prepared by Reed4 have silylium ion
character or not. The ®rst `free' silylium ion was reported
by Lambert and Zhao in 1997, that is, the trimesitylsilylium
ion stabilized mainly by the steric protection of the silicon
atom by bulky substituents.12 Also a cation having a non-
classical delocalized structure was shown to be free from
coordination with solvents or counter anion.13 More
recently a salt of a homocyclotrisilenylium ion was isolated
by Sekiguchi and was found to have no interaction with
counter anion from the X-ray structural analysis.14 In this
case, the aromatic homoconjugation as observed in the
carbon analog is supposed to stabilize the silylium ion
center.
2
As another potentially aromatic system, a silicon analog of
the tropylium ion *1), i.e. the silatropylium ion *2), is also a
good candidate for the examination of the above question.
However, the attempted preparation of the dibenzosilatro-
pylium ion *3) by Olah's group was not successful and they
concluded that silylium ions are not stabilized in solution by
incorporation of the silicon atom into a potentially aromatic
ring system.5b Even in the gas phase, a recent suggestion of
the silatropylium ion *2) as one of the C6H7Si1 isomers
observed in an FT ICR mass spectrum22 was disproved
by a subsequent experimental study showing that it was a
rearranged isomer, C6H6´SiH1.23 Theoretical studies have
Keywords: aromaticity; cycloheptatrienes; NMR; silicon heterocycles;
theoretical studies; tropylium ions; X-ray crystal structures.
Corresponding author. Tel.: 181-774-38-3172; fax: 181-774-38-3178;
e-mail: komatsu@scl.kyoto-u.ac.jp
p
0040±4020/01/$ - see front matter q 2001Elsevier Science Ltd. All rights reserved.
PII: S0040-4020*01)00253-8

3646
T. Nishinaga et al. / Tetrahedron 57 '2001) 3645±3656
Scheme 1.
indicated that cation 2 is less stable than `silabenzyl cation'
by 9 kcal mol21 24
Thus the synthesis and characterization
of the silatropylium ion remains as a major challenge in
organosilicon chemistry.
substituents and *2) the cation should be prepared in the
presence of a low-coordinating counter anion and *3) in a
low coordinating solvent.1e,f,12 In addition, the annealation
with benzene ring*s) is known to thermodynamically
destabilize the tropylium ion *1),25 and therefore would
not be appropriate for the study of the silatropylium ion.
For the realization of the silatropylium ion, we designed
a modi®cation of the seven-membered ring with rigid
s-frameworks, that is, annealation with bicyclo[2.2.2]-
octene *abbreviated as BCO) units. In our previous study,
such structural modi®cation with BCO units was found to be
the most effective in stabilizing the tropylium ion.26 In the
present paper are described the detailed accounts on the
preparation and characterization of the ®rst silatropylium
ion derivative 5 stabilized by such a structural modi®cation
and the presence of a bulky mesityl group on silicon;27 also
an attempt made to prepare a derivative with a methyl group
on silicon *4) and the reactions of 4 and 5 with acetonitrile
or perchlorate ion are described together with the results of
theoretical calculations.
.
The attempted observation of cation 3 was performed in
acetonitrile with perchlorate as a counter anion5b before a
recent guideline was established for the possible generation
of silylium ion in condensed phase. The guideline is that *1)
the silylium ion center should be surrounded by bulky
2. Results and discussion
2.1. Preparation and characterization of the
silatropylium ion 5
For the preparation of silatropylium ions 4 and 5, the precur-
sor silacycloheptatrienes *silepins) 728 and 10 were prepared
from the dibromide of BCO trimer 629 in the following way,
as shown in Scheme 1. Dibromide 6 was dilithiated with
4 equiv. of t-BuLi at 2608C and then allowed to react
with dichloromethylsilane or with tetrachlorosilane to give
silepin 7 in 66% or dichlorosilepin 828 in 55% yield, respec-
tively. Dichlorosilepin 8 was allowed to react with mesityl-
lithium to give mesitylchlorosilepin 9 in 40% yield, which
was then reduced to silepin 10 in 90% yield by treatment
with LiAlH4. The molecular structures of silepins 9 and 10
determined by X-ray crystallography are shown in Fig. 1. In
Figure 1. ORTEP drawings *50% probability) showing the X-ray crystal
structures for 9 *a) and 10 *b).

T. Nishinaga et al. / Tetrahedron 57 '2001) 3645±3656
3647
²
to silatropylium ion 5, as shown in Fig. 2 and Table 1. Here,
apparently a bulky substituent such as a mesityl group was
necessary for stabilizing the silatropylium ion. At tempera-
³
tures higher than 2508C, silatropylium ion 5 decomposed to
give benzene 11 as was observed in C7D8 at low tempera-
tures. Thus, CD2Cl2 with its medium polarity and small size
is favorable for approaching the silylium ion center to
weakly coordinate to it and preventing the rearrangement
at the temperature below 2508C while C7D8 with its lower
polarity and a larger molecular size cannot stabilize the
cationic species by coordination9,11 even at low temperatures.
The 29Si NMR signal of silatropylium ion 5 was observed at
d 142.9 ppm in CD2Cl2, which is 192.2 ppm down®eld
shifted compared with the precursor silepin 10 *d 249.3).
This is taken as clear evidence for the silylium ion character
of 5. The observed 29Si chemical shift is also in fair agree-
ment with the value calculated for 5 *d 159.9; GIAO/HF/6-
3111G*2df,p)*Si), 6-311Gp//B3LYP/6-31Gp).30 This
method of calculation is known to reproduce the experi-
mental value observed for the trimesitylsilylium ion.31 The
29Si chemical shift of 5 is 28 ppm down®eld shifted
compared with the solid-state 29Si NMR chemical shift *d
115) of i-Pr3Si1CB11H6Cl62, which was shown to have
some interaction between the silylium ion center and the
chlorine atom of the counter ion.4d Thus, the interaction
Scheme 2.
the crystal, the mesityl group takes the `axial' position for 9
and the `equatorial' position for 10. The molecular structure
of 10 with the `equatorial' mesityl group might be unfavor-
able for the hydride abstraction by such a bulky reagent like
triphenylmethyl cation for steric reasons. In solution,
however, the rapid ring inversion should take place28 to
bring the mesityl group to the axial position for the hydride
abstraction from the Si±H bond to proceed readily.
First the hydride abstraction was attempted in the presence
of a non-coordinating anion in a non-coordinating solvent.
When an attempt was made to generate the silatropylium ion
4 or 5 by hydride abstraction from the silepin 7 or 10 with an
equivalent of triphenylmethyl tetrakis*penta¯uorophenyl)-
borate *TPFPB) in toluene-d8 *C7D8) under vacuum at
between the silylium ion center of 5 and CD2Cl2 should
.
2
be smaller than the case for i-Pr3Si1CB11H6Cl6
The chemical shift of the BCO bridgehead *bh) protons is
useful for judging the presence of a diamagnetic ring current
because they are rigidly ®xed in the same plane as a
silatropylium ion ring.32 As shown in Fig. 2b, the signals
for these protons of 5 were observed at d 3.70, 3.64, and
3.21ppm. These signals except for the signal at d 3.21were
0.8±0.9 ppm down®eld shifted compared with those for the
neutral precursor 10 *dbh 2.81*4H), 2.61*2H)), indicating
the presence of a diamagnetic ring current and thus aroma-
ticity in 5. The signal at d 3.21is assigned to the proton *H a)
which sticks out in the shielding zone of a mesitylene ring.
This up®eld shift of ca. 0.4 ppm due to the mesityl group
was also reproduced by the calculations *dbh 3.2, 3.2, 2.8;
GIAO/HF/6-3111G*2df,p)*Si), 6-311Gp//B3LYP/6-31Gp).
1
2508C, only a ring contraction occurred to show the H
NMR signals for the BCO-annealated benzene 11 *Scheme
2), possibly via a non-classical ion, 11´R±Si1, which corre-
sponds to C6H6´SiH1 observed in the gas phase.23 On the
other hand, when the hydride abstraction was conducted in
dichloromethane-d2 *CD2Cl2) at 2508C, the reaction of sile-
pin 10 gave the 29Si, 1H and 13C NMR signals corresponding
²
At the hydride abstraction reaction, ca. 0.5 equiv. of trityl cation was not
consumed and about 6% of 11 and mesitylene were also formed. The fate
of the possible species such as Mes±Si1 generated upon the ring contrac-
tion of 5 is not known, but such a species and its degradation products
would also work to abstract hydride from 10.
³
When the hydride abstraction of the methyl-substituted silepin 7 was
conducted in CD2Cl2, the NMR measurements at room temperature *1H
NMR d 3.54 *s, 6H), 2.03 *d, 12H), 1.62 *d, 6H), 1.28 *d, 6H): 13C NMR d
34.5, 26.9, 24.8; the signal for the sp2 carbon could not be identi®ed)
revealed that the product was not the silatropylium ion 4 but presumably
the rearranged isomer with a non-classical structure, i.e. 11´Me±Si1 or
11´Me±SiCl21. Further characterization will be reported elsewhere.
Figure 2. *a) 29Si NMR *72.5 MHz) and *b) 1H NMR spectra *400 MHz) for
5 at 2508C in CD2Cl2. The signals marked with 1, p, and # correspond to
those for 11, mesitylene, and triphenylmethane, respectively.

3648
T. Nishinaga et al. / Tetrahedron 57 '2001) 3645±3656
Table 1. The 1H, 13C, and 29Si NMR chemical shifts *d, ppm) for 5, 7, 10, and 16±19 in CD2Cl2
Compd
1H NMR
CH
13C NMR
29Si NMR
142.9
Mes-H
SiH
CH2
CH3
Tropyl
Aryl
CH
CH2
CH3
CN
a
5
±
3.70
*2H)
3.64
*2H)
3.21
*2H)
2.80
*4H)
2.74
*2H)
1.85
2.41
*6H)
2.36
*3H)
175.9
153.2
149.7
144.8
143.8
128.0
118.7
36.8
35.1
34.9
25.1
24.2
23.9
25.9
21.4
*12H)
1.30
*12H)
7b
4.47
*1H)
1.7±1.0
*24H)
0.22
*3H)
151.2
141.4
134.5
34.2
33.9
33.0
26.8
26.7
26.5
26.3
25.8
25.7
27.1
26.7
26.0
25.9
25.2
25.9
25.3
25.1
25.0
24.6
24.0
28.2
25.7
25.2
24.6
23.1
20.6
±
c
10d
6.78
*2H)
4.81
*1H)
2.81
*4H)
2.61
*2H)
1.8±1.0
*24H)
2.40
*6H)
2.22
*3H)
152.0
145.5
141.8
138.7
128.2
127.9
34.4
33.5
33.1
24.6
20.8
249.3
21.7
16d,e
2.95
*2H)
2.90
*2H)
2.74
*2H)
2.85
*2H)
2.76
*2H)
2.65
*2H)
2.93
*2H)
1.8±0.9
*24H)
2.61
*3H)
0.83
*3H)
159.1
141.1
130.0
32.6
31.3
25.9
120.4
125.6
17d,e
6.70
*2H)
1.8±1.0
*18H)
1.0±0.6
*4H)
0.17
*2H)
3.07
*3H)
2.25
*6H)
2.13
*3H)
0.80
*3H)
153.6
145.8
141.4
142.4
130.1
128.8
118.7
32.4
32.3
31.7
22.5
20.8
5.4
215.9
18
1.8±0.9
*24H)
157.2
141.1
33.5
33.4
26.6
26.1
25.7
8.9
35.4
322.4.911 25.5
*2H)
2.85
25.3
25.2
24.5
28.5
26.6
26.0
25.2
25.1
24.9
*2H)
19
6.71
*2H)
3.10
*2H)
2.74
*2H)
2.67
*2H)
1.8±1.2
2.36
*6H)
2.20
*3H)
151.6
146.5
141.5
141.3
135.4
128.9
123.0
33.0
32.9
31.6
24.3
20.9
26.0
*18H)
1.1±0.7
*4H)
0.59
*2H)
a
A signal for the two mesityl ring protons is overlapped with signals for triphenylmethane.
In C6D6.
The signal could not be observed.
There is some overlap in 13C NMR signals.
At 2808C.
b
c
d
e
The 13C NMR signals for the sp2 carbons of 5 in CD2Cl2 *d
2´CH2Cl2, and 5´CH2Cl2. As shown in Fig. 3a and b, while 2
took a completely planar conformation, the central seven-
membered ring of the optimized structure for 5 at B3LYP/6-
31Gp level was not planar but in a boat form. This non-
planarity of the seven-membered ring in 5 is apparently
due to the steric repulsion between bridgehead protons at
the stern part. The bending angles at the bow and the stern
parts were 128 and 278, respectively. Like the other struc-
tural features of 5, the three carbon atoms around the silicon
atom were located in almost the same plane with silicon *the
sum of the bond angles, 3598) and the mesitylene ring was
perpendicular to the mean plane of the silatropylium ion
ring.
175.9, 153.2, 149.7 ppm; Table 1) showed down®eld shifts
compared with the signals of silepin 10 in CD2Cl2 *d 152.0,
145.5, 141.8), indicating considerable positive-charge
delocalization in the seven-membered ring. The most down-
®eld shifted signal *d 175.9) is assigned to the b-carbon
based on the calculated chemical shifts for 5 *a-C,
147 ppm; b-C, 185; g-C, 150). These results imply that a
contribution of the sila-allylic resonance structure is import-
ant for the positive-charge delocalization.
In order to estimate the structure and aromaticity of the
silatropylium ion and the degree of interaction between
the silatropylium ion and CH2Cl2, theoretical calculations
were performed for Me3Si1, tropylium ion *1), silatropyl-
ium ion *2), and BCO-annealated silatropylium ion 5,
together with the CH2Cl2 complexes, i.e. Me3Si1´CH2Cl2,
In spite of the non-planarity of the central ring in 5, the nucleus
independent chemical shifts21b at 1A above and below the
Ê
ring *NICS*1); GIAO/HF/6-311Gp//B3LYP/6-31Gp) for 5

T. Nishinaga et al. / Tetrahedron 57 '2001) 3645±3656
3649
Figure 3. The optimized structures of 2 *a), 5 *b), Me3Si1´CH2Cl2 *c), 2´CH2Cl2 *d), and 5´CH2Cl2 *e) at B3LYP/6-31G*d) level. The Si±Cl distances and
Mulliken charges on silicon atoms and other units are given.
were 26.5 and 27.5, respectively, which are comparable to
those of 2 *NICS*1)ˆ27.3). This result supported the
conclusion that the observed down®eld shifts of bridgehead
protons for 5 were due to a diamagnetic ring current on a
silatropylium ion ring. For the estimation of the aromaticity
of 5, the down®eld shifts of the bridgehead protons were
compared with those for the carbon analog, 12. The down-
®eld shifts for 12 upon going from the neutral precursor *dbh
2.88 *2H), 2.67 *4H)) to the ion *dbh 4.09, 4.06, 3.17)26c
were 1.2±1.4 ppm except for Ha. Thus, the down®eld shifts
in 5 *0.8±0.9 ppm except for Ha) were somewhat smaller
than the values in 12, probably due to the less effective
aromatic conjugation between the 3p orbital of silicon and
2p orbitals of carbons, but the aromaticity of 5 is com-
parable to that of 2 as mentioned above. Based on the theo-
retical indices for aromaticity, i.e. NICS*1) and magnetic
susceptibility exaltation *L)33 for unsubstituted tropylium
ion *1) and silatropylium ion *2), the aromaticity of 2
*NICS*1)ˆ27.3, Lˆ215.6) could be roughly estimated
to be about 70% of that of 1 *NICS*1)ˆ210.7, Lˆ220.1).
As concerns the subject of the interaction between the
silylium ion center and the solvent, CH2Cl2, the Si±Cl
distance for 2´CH2Cl2 *Fig. 3d) was calculated to be consid-
Ê
Ê
erably longer *2.735 A) than that for Me3Si±Cl *2.112 A) as
a model compound for the covalent molecule. This Si±Cl
Ê
distance for 2´CH2Cl2 is still 0.33 A longer than that for
1
Me3Si ´CH2Cl2 *2.407 A; Fig. 3c), and also the Mulliken
Ê
charge on CH2Cl2 in 2´CH2Cl2 *10.17) is less than that in
Me3Si1´CH2Cl2 *10.28). Furthermore, the stabilization
energy for CH2Cl2 coordination in 2´CH2Cl2 *7.7 kcal
mol21) as estimated from the energy difference between
2´CH2Cl2 and the sum of 2 and CH2Cl2 is smaller than the
case for Me3Si1´CH2Cl2 *18.2 kcal mol21). Thus, the cyclic
conjugation of the silatropylium ion apparently reduces the
interaction between the silylium ion center and CH2Cl2.
In addition, the optimized structure of 5´CH2Cl2 showed that
Ê
the Si±Cl distance *5.390 A) was further elongated. The
Si±Cl distance is much longer than the sum of the van der
Ê
Waals radii of silicon and chlorine *3.8±4.0 A), indicating
no direct interaction between the silylium center and
CH2Cl2. The absence of the interaction would be ascribed
not only to the bulky substituent on the silicon atom but also
to the thermodynamic stability of 5 caused by annelation
with BCO units. From the comparison of the charge on
silicon of 2 *10.56) and 5 *10.39), BCO units are shown

3650
T. Nishinaga et al. / Tetrahedron 57 '2001) 3645±3656
Scheme 3.
to be effective in further delocalizing the positive charge on
silicon. On the other hand, the stabilization energy esti-
mated from the energy difference between 5´CH2Cl2 and
the sum of 5 and CH2Cl2 was not negligible *2.5 kcal
mol21). This stabilization energy would be caused mainly
by the dipole interaction between 5 and the CH2Cl2 mol-
ecule and would be important for the stabilization of 5 in
solution.
temperature under inert atmosphere: even the methyl sub-
stituted derivative 16 did not undergo the rearrangement as
was observed in CD2Cl2 in the absence of CH3CN.
The silatropylium ion 5 is stable at temperatures below
2508C, but it slowly abstracts the chloride ion from
CD2Cl2 to furnish dichlorosilepin 8 at 2508C *t1/2,6 h).
The possible mechanism is illustrated in Scheme 3. First,
the chloride ion is abstracted slowly from the solvent by 5 to
produce silepin 9 and the chloromethyl cation. Such an
abstraction of the chloride ion from CH2Cl2 by the silylium
ion has been reported previously.4a,34 Then the produced
chloromethyl cation exerts an electrophilic aromatic sub-
stitution on the mesitylene ring of 9 to cleave the Si±Mes
bond to generate the highly reactive chlorosilatropylium ion
13 and *chloromethyl)mesitylene 14. The second chloride
abstraction by 13 from 14 furnishes dichlorosilepin 8 and
the benzyl cation 15. Finally, the benzyl cation 15 abstracts
hydride from triphenylmethane which was generated upon
the formation of 5. In support of the above mechanism, the
formation of tetramethylbenzene-d2 *C10H12D2) was
con®rmed by GC-MS analysis.
As shown in Fig. 4, the optimized structure of 17 at the
B3LYP/6-31G*d) level demonstrates that a CH3CN mol-
ecule is strongly coordinated to the silicon atom of 5, in
contrast to the case of 5´CH2Cl2, due to the much greater
nucleophilicity of CH3CN. Also the calculated Mulliken
charge on total BCO units *10.10) supports the conclusion
derived from the NMR results described above that the
positive charge is not substantially delocalized into
the seven-membered ring carbons. On the other hand, the
Mulliken charge of CH3CN in 17 *10.29) and the stabili-
zation energy for CH3CN coordination on 5 *14.2 kcal
mol21), estimated from the energy difference between 17
and the sum of 5 and CH3CN, were smaller than the cases
for Me3Si1´CH3CN *10.36, 51.3 kcal mol21) and 2´CH3CN
*10.36, 34.6 kcal mol21). Also the distance of Si±N
Ê
Ê
Ê
*1.926 A) for 17 was 0.03 A and 0.01A longer than the
Ê
calculated Si±N distance for Me3Si´CH3CN *1.894 A) and
Ê
for 2´CH3CN *1.916 A), respectively. These results indicate
that the interaction of the BCO-annelated silatropylium ion
2.2. Reaction of the silatropylium ions with acetonitrile
When the hydride abstraction of silepin 7 or 10 using tri-
phenylmethyl TPFPB was conducted in CD2Cl2 in the
1
presence of 2 equiv. of CH3CN, the 29Si, H and 13C NMR
signals corresponding to the CH3CN complex, 16 or 17,
were observed *Table 1). Upon comparison of the NMR
data of 17 with the silatropylium ion 5 and with the pre-
cursor silepin 10, the chemical shifts of 17 were closer to
those of 10 rather than 5. The 29Si NMR signal of 17 showed
only 33.4 ppm down®eld shift from that of 10. The bridge-
head protons and the sp2 carbons also showed only a slight
down®eld shift from those of 10. These results indicate that
the diamagnetic ring current in 5 was quenched by coordi-
nation with CH3CN and the positive charge is delocalized
not into the seven-membered ring but into CH3CN. The
complexes 16 and 17 were found to be stable at room
Figure 4. The optimized structure of 17 at B3LYP/6-31G*d) level. The Si±
N distance and Mulliken charges on silicon atoms and other units are given.

T. Nishinaga et al. / Tetrahedron 57 '2001) 3645±3656
3651
Figure 5. The expanded 1H NMR spectra and their simulations *a)±*c) for 16 and *d)±*f) for 17 at various temperatures. The signals with arrows are for the
methyl groups of complexed and free acetonitrile.
5 with CH3CN is somewhat reduced due to the presence of
BCO units and the mesityl group. However, the combi-
nation of these effects was not suf®cient for making the
silylium ion 5 free from the interaction with CH3CN.
our dynamic NMR experiment on 16 and 17 showed the
opposite result; the transition state involves the association
process, probably via a pentacoordinated silicon.
In the 1H NMR for 17 at 2208C *Fig. 5f), two sharp signals
for the free and complexed CH3CN were observed with the
integrated ratio of 1:1. This shows that 1 equiv. of CH3CN
was coordinated to 5 forming a silylium ion±CH3CN 1:1
complex as shown in the above calculations and not a 1:2
complex having a pentavalent silicon. As the temperature
was raised *Fig. 5d and e), broadening of the signals was
observed, indicating that the free and complexed CH3CN
molecules were exchanging rapidly on an NMR time
scale. In this system, however, the coalescence point
could not be reached due to the low boiling point of the
solvent *CD2Cl2). On the other hand, in the case of 16, the
independent signals for free and complexed CH3CN, which
were observed at low temperatures *e.g. 2908C; Fig. 5c),
were coalesced and became a single signal at room tem-
perature. From the line shape analyses and Arrhenius
For the CH3CN complexes 16 and 17, the exchange between
the free and complexed CH3CN was found to be observable
with the variable temperature 1H NMR technique as shown
in Fig. 5. Such dynamic behavior of the CH3CN exchange
on silylium ion in solution is of interest in connection with
the mechanism of the reaction involving a recently reported
chiral and Lewis-acidic silyl cationic catalyst having a 1,10-
binaphtyl backbone.35 A claim for the silyl cation as the
catalyst was contradicted by Olah et al., who concluded
that the catalyst is not the silylium ion but the silylnitrilium
ion.36 Once the transition state for the exchange of CH3CN
in a trialkylsilylated acetonitrilium ion was suggested to
involve the dissociation process forming free CH3CN and
the silylium ion.37 If this is the case, the catalyst would also
involve the silylium ion state. However, as described below,
Scheme 4.

3652
T. Nishinaga et al. / Tetrahedron 57 '2001) 3645±3656
plots, activation parameters for the exchange process were
³
³
determined to be DH ˆ5.7^0.5 kcal mol21 and DS ˆ
219.5^2 eu for 16 and DH ˆ7.3^0.5 kcal mol21 and
³
³
DS ˆ225.0^2 eu for 17. The large negative values of
³
DS clearly indicate that the transition-state structure is
highly ordered with much steric constraint for both cases
and should involve the pentacoordinated silicon as shown in
Scheme 4.
The formation of a pentacoordinated silicon at the transition
state would be caused by the high Lewis acidity of the
silicon atom in silylnitrilium ions 16 and 17. Such a penta-
valent silyl cation coordinated with two CH3CN molecules
was previously suggested to form in an equilibrium
with tetracoordinated species when an excess amount of
CH3CN was present.7 In addition, the dissociation of the
complex into CH3CN and silylium ion requires more
energy for the non-p-conjugated silylium ion than for
the p-conjugated one *see above). Thus, the mechanism
of the CH3CN exchange for the silylium ion with non-
p-conjugation would also not involve the trivalent silylium
ion state but the pentavalent state, and hence the
catalyst having a binaphthyl backbone would be best
described as the silylnitrilium ion and not the silylium ion,
as pointed out by Olah.36
Figure 6. The optimized structure of 19 at B3LYP/6-31G*d) level. The Si±
O distance and Mulliken charges on silicon atoms and other units are given.
mol21). Thus, the cyclic conjugation and steric congestion
in 5 again reduced the interaction between the silylium ion
and perchlorate ion, but the combination of these effects was
not suf®cient for making the silylium ion free from the
interaction with the perchlorate ion.
The perchlorate ester 18 was found to react with water
causing a novel rearrangement with ring contraction to
1
give a new polycyclic hydrocarbon. Based on H and 13C
NMR spectra with DEPT measurements, mass spectrum,
and elemental analysis, the product *60% yield) was
assumed to have a cyclopentene structure with a spiro-
connected bicyclic framework. The possible structure of
this product appeared to be either the isomer 20a or 20b,
and the NMR analysis revealed that the product was
composed of a single isomer. Its structure was ®nally deter-
mined by X-ray crystallography to be 20a, as shown in Fig. 7.
2.3. Reaction of the silatropylium ions with perchlorate
ion
In contrast to the ionic salt of triphenylmethyl perchlorate,
triphenylsilylium ion reacts with the perchlorate ion to give
the covalent perchlorate ester.5a In order to examine again if
the combination of aromatic stabilization and steric pro-
tection in silylium ion 5 prevents the formation of such a
covalent bond, the hydride abstraction from silepins 7 and
10 was conducted using triphenylmethyl perchlorate at low
temperatures to room temperature. For both reactions, the
1
29Si, H and 13C NMR signals of the products 18 and 19
*Table 1) showed only small down®eld shifts from the
neutral precursors 7 and 10 as in the cases of CH3CN
complexes 16 and 17, indicating that the silatropylium ion
character was lost and covalent perchlorate esters were
formed. These esters 18 and 19 were also stable at room
temperature under inert atmosphere.
The possible mechanism for the formation of 20a is
depicted in Scheme 5. First, water attacks the highly
Lewis acidic silicon atom of 18 to give hydroxysilepin
and perchloric acid, which protonates the triene part to
form the Int-1. The PM3 calculations indicated that the
Int-1would rearrange to the Int-2, in which a double bond
is strongly coordinated to the silyl cation with the carbon
atoms charged negatively. Then the second attack of water
As shown in Fig. 6, the optimized structure of ester 19 at the
B3LYP/6-31G*d) level indicates that the distance of Si±O is
Ê
Ê
1.787 A, which is 0.105 A longer than that for Me3SiOH
Ê
*1.682 A) as a model compound for the covalent compound.
Ê
Also, this distance is 0.021and 0.027 A longer than the
Ê
calculated Si±O distance for Me3SiClO4 *1.766 A) and for
Ê
2´ClO4 *1.760 A), respectively. The stabilization energy for
the formation of a covalent bond between silatropylium ion
5 and the perchlorate ion *104.7 kcal mol21), estimated
from the energy difference between 19 and the sum of 5
and the perchlorate ion, was smaller than the cases for
Me3Si´ClO4 *152.1 kcal mol21) and 2´ClO4 *133.7 kcal
Figure 7. ORTEP drawing *50% probability) showing the X-ray crystal
structure of 20a.

T. Nishinaga et al. / Tetrahedron 57 '2001) 3645±3656
3653
Scheme 5.
molecule occurs to give the Int-3, which has the pentadienyl
cation p-system. The thermal electrocyclization of the Int-3
proceeds in a conrotatory fashion, thus establishing the
stereoselectivity at the spiro-junction carbon in the Int-4.
For the moment we do not have good reasoning for the
following process, but apparently two hydrogen atoms are
required to replace the siliconium ion in the Int-5 to furnish
hydrocarbon 20a.
Thus, although these compounds did not show any silylium
ion character, they still have unique properties which were
not observed in the neutral precursors, silepins 7±10.
4. Experimental
4.1. General
Melting points were measured on a Yanaco MP-500D appa-
ratus and are uncorrected. Elemental analyses were
performed at the Microanalysis Division of Institute for
3. Conclusion
1
Chemical Research, Kyoto University. H, 13C and 29Si
We have succeeded in the ®rst NMR observation of the
silatropylium ion derivative 5 with the aid of annealation
with rigid BCO frameworks and of placement of a bulky
substituent on silicon. This study revealed not only that the
silatropylium ion is aromatic but that the aromatic stabili-
zation is effective for reducing the interaction between the
silylium ion center and a solvent having weak nucleo-
philicity like CH2Cl2. On the other hand, when a solvent
with stronger nucleophilicity such as CH3CN was added
or the perchlorate ion was present instead of TPFPB as a
counter anion, the combination of the aromatic stabilization
and steric protection could not prevent the formation of a
covalent bond between the silylium ion center and CH3CN
or the perchlorate ion.
NMR spectra were recorded on a Varian Mercury-300
*1H, 13C) or a JEOL JNM-GX 400 *1H, 13C, 29Si) spectro-
meter. GC-MS spectrum was measured on a Shimadzu QP
5050A system. Preparative GPC separation was performed
with a JAI LC-908 chromatograph equipped with JAIGEL
1H and 2H columns using toluene as eluent. THF was
distilled over sodium benzophenone ketyl. Dibromide of
BCO trimer *6) was prepared following the literature pro-
cedure.29 All the reagents used were commercial materials
unless otherwise noted.
4.1.1. 1-Hydro-1-methyl-2,3:4,5:6,7-tris2bicyclo[2.2.2]-
octeno)silacycloheptatriene 27). To a stirred solution of
dibromide 6 *210 mg, 0.44 mmol) in THF *10 mL) at
2608C was added dropwise a solution of 1.47N t-BuLi in
pentane *1.2 mL, 1.76 mmol). Addition of each drop of
t-BuLi solution caused yellow coloration. After stirring
for 15 min at 2608C, dichloromethylsilane *0.08 mL,
88 mg, 0.77 mmol) was added dropwise to the reaction
mixture, which gradually turned colorless. After being
stirred for 30 min at 2608C, the mixture was allowed to
warm to room temperature. The solvent was then removed
by evaporation and the crude product was extracted with
hexane. The extract was puri®ed by preparative GPC eluted
with toluene to give 7 *105 mg, 66%) as a white solid. 7: mp
.3008C *dec). For NMR data, see Table 1. Anal. Calcd for
C25H34Si: C, 82.80; H, 9.45. Found: C, 83.00; H, 9.43.
For the CH3CN complexes 16 and 17, the exchange process
of CH3CN could be observed with the dynamic NMR tech-
nique. Such a measurement on the exchange of CH3CN on a
silylium ion was performed for the ®rst time. The exchange
process was found to proceed through a highly ordered
transition state, which would involve the pentacoordinated
silicon atom. On the other hand, the reaction of perchlorate
ester 18 with water caused a novel rearrangement with a
ring contraction to give hydrocarbon 20a. This novel reac-
tion is considered to be triggered by the attack of water on
silicon atom of 18. In conclusion, both the CH3CN exchange
reaction in 16 and 17 and the rearrangement observed for 18
are due to the high Lewis acidity of silicon atom in 16±18.

3654
T. Nishinaga et al. / Tetrahedron 57 '2001) 3645±3656
4.1.2. 1,1-Dichloro-2,3:4,5:6,7-tris2bicyclo[2.2.2]octeno)-
silacycloheptatriene 28). To a stirred solution of dibromide
6 *204 mg, 0.43 mmol) in THF *10 mL) at 2508C was
added dropwise a solution of 1.5N t-BuLi in pentane
*1.14 mL, 1.7 mmol). After stirring for 15 min at 2608C,
tetrachlorosilane *0.06 mL, 89 mg, 0.52 mmol) was added
dropwise to the reaction mixture, which gradually turned
colorless. After being stirred for 30 min at 2508C, the
mixture was allowed to warm to room temperature. The
solvent was then removed by evaporation and the crude
product was extracted with hexane. The extract was puri®ed
by preparative GPC eluted with toluene to give 8 *98 mg,
55%) as a white solid. 8: mp 194±1958C; 1H NMR
*300 MHz, C6D6) d 3.30 *s, 2H), 2.78 *s, 2H), 2.70 *s,
2H), 1.1±1.5 *m, 24H); 13C NMR *67.8 MHz, C6D6, 258C)
d 155.1, 141.8, 139.3, 33.7, 32.4, 32.3, 26.3, 25.9, 25.2.
Anal. Calcd for C24H30Cl2Si: C, 69.05; H, 7.24. Found: C,
68.53; H, 7.37.
*0.75 mL), dried with CaH2 overnight and with P2O5 for
1h, was transferred into the Pyrex glass tube cooled with
liquid nitrogen under vacuum, and the glass tube was sealed
off. The reaction mixture was stirred at 2508C for 0.5 h.
Then the solution was transferred into the side arm at
2788C. The cooled NMR tube was sealed off and was
subjected to the NMR measurements.
4.1.6. Preparation of acetonitrile complexes 16 and 17 in
CD2Cl2. Into a 5 mm diameter NMR tube which was
connectable to a vacuum line, 7 or 10 *0.050 mmol) and
triphenylmethyl tetrakis*penta¯uorophenyl)borate *46 mg,
0.050 mmol) were placed, and the tube was connected to
a vacuum line. CD2Cl2 *0.75 mL) and two equivalents of
CH3CN *5.2 mL, 0.10 mmol), dried with CaH2 overnight
and with P2O5 for 1h, was transferred into the NMR tube
cooled with liquid nitrogen under vacuum. The cooled
NMR tube was sealed off, shaken well, and subjected to
the NMR measurements.
4.1.3. 1-Chloro-1-mesityl-2,3:4,5:6,7-tris2bicyclo[2.2.2]-
octeno)silacycloheptatriene 29). A solution of mesityl-
lithium *0.18 mmol), prepared from bromomesitylene in
THF and two equivalents of t-butyllithium in pentane, was
added to a stirred solution of dichlorosilepin 8 *50 mg,
0.12 mmol) in THF *3 mL). The solution was stirred for
0.5 h at room temperature. The volatile compounds were
removed in vacuo and the resulting pale yellow mixture
was extracted with hexane and the insoluble materials
were removed by ®ltration. The ®ltrate was subjected to
preparative GPC eluted with toluene. Separation was
effected by recycling the crude products eight times to
give 9 *24 mg, 40%) as a white solid. A single crystal was
obtained by slow recrystallization from pentane. 9: mp 167±
1688C. 1H NMR *400 MHz, CD2Cl2) d 6.64 *s, 2H), 3.17 *t,
2H), 2.72 *t, 2H), 2.63 *s, 2H), 2.31*s, 6H), 2.17 *s, 3H),
1.8±1.4 *m, 16H), 1.18 *d, 2H), 0.94 *m, 2H), 0.76 *m, 2H),
0.31*d, 2H); 13C NMR *100 MHz, CD2Cl2) d 149.5, 144.7,
141.2, 139.7, 137.7, 129.6, 128.7, 33.5, 33.4, 32.0, 29.1,
27.1, 26.6, 25.8, 25.6, 24.7, 21.1; 29Si NMR *78.5 MHz,
CD2Cl2) d 211.2. Anal. Calcd for C33H41ClSi: C, 79.08;
H, 8.25. Found: C, 79.11; H, 8.26.
4.1.7. Preparation of perchlorates 18 and 19 in CD2Cl2.
Into a 5 mm diameter NMR tube which was connectable to a
vacuum line, 7 or 10 *0.050 mmol) and triphenylmethyl
perchlorate *17 mg, 0.050 mmol) were placed, and the
tube was connected to a vacuum line. CD2Cl2 *0.75 mL),
dried with CaH2 overnight and with P2O5 for 1h, was trans-
ferred into the NMR tube cooled with liquid nitrogen under
vacuum. The cooled NMR tube was sealed off, shaken well,
and subjected to the NMR measurements.
4.1.8. Reaction of perchlorate 18 with water. Into a stirred
CD2Cl2 solution *0.75 mL) of 18 generated from 7 *9.6 mg,
0.026 mmol) and triphenylmethyl perchlorate *9.1mg,
0.026 mmol), 10 mL *0.56 mmol) of water was added via
a microsyringe. After a few minutes, the reaction mixture
was evaporated in vacuo and separated with preparative
HPLC *SiO2) eluted with hexane to give 20a *5.0 mg,
60% yield) as a white solid. A single crystal was obtained
by slow recrystallization from pentane. 20a: 1H NMR
*300 MHz, C6D6) d 2.79 *dd, 1H), 2.73 *q, 1H), 2.48 *q,
1H), 2.28 *d, 1H), 1.9±1.1 *m, 30H); 13C NMR *100 MHz,
C6D6) d 150.7 *C), 143.4 *C), 53.0 *C), 49.6 *CH), 48.2
*CH), 37.3 *CH), 34.3 *CH2), 31.5 *CH), 30.1 *CH), 29.6
*CH2), 28.7 *CH2), 27.6 *CH), 27.3 *CH2), 27.1*CH), 26.8
*CH2), 26.5 *CH2), 26.1*CH 2), 25.9 *CH), 25.9 *CH2), 25.8
*CH2), 25.0 *CH2), 24.3 *CH2), 22.9 *CH2), 21.5 *CH2).
Anal. Calcd for C24H34: C, 89.37; H, 10.63. Found: C,
89.29; H, 10.66.
4.1.4. 1-Hydro-1-mesityl-2,3:4,5:6,7-tris2bicyclo[2.2.2]-
octeno)silacycloheptatriene 210). A THF solution of
LiAlH4 *1M, 0.250 mL, 0.25 mmol) was added to
9
*50 mg, 0.10 mmol) in THF *1.5 mL). The reaction mixture
was stirred at 408C for 18 h. The volatile compounds were
removed in vacuo and the resulting mixture was extracted
with hexane and the insoluble materials were removed by
®ltration. The ®ltrate was subjected to preparative GPC
eluted with toluene to give 10 *42 mg, 90%) as a white
solid. A single crystal was obtained by slow recrystalli-
zation from pentane. 10: mp 1 858C *dec). For NMR data,
see Table 1. Anal. Calcd for C33H42Si: C, 84.91; H, 9.07.
Found: C, 84.52; H, 9.29.
4.1.9. X-Ray crystallography of 9, 10, and 20. Data for
compounds 9 and 10 were collected on a Rigaku AFC7R
with graphite monochromated CuKa radiation and data for
20 were collected on a Bruker SMART APEX with graphite
monochromated MoKa radiation.
Crystal data for 9 are as follows: C33H41ClSi; space group
Ê
Ê
Ê
4.1.5. Preparation of silatropylium ion 5 in CD2Cl2. A
10 mm diameter Pyrex glass tube, which was connectable
to a vacuum line and equipped with a 5 mm diameter NMR
tube as a side arm, was prepared. Into the Pyrex glass tube
were placed 10 *15.4 mg, 0.0330 mmol) and triphenylmethyl
tetrakis*penta¯uorophenyl)borate38 *30.8 mg, 0.0330 mmol),
and the tube was connected to a vacuum line. CD2Cl2
P-1; aˆ15.584*2) A, bˆ17.833*3) A, cˆ9.974*1) A, bˆ
Ê 3
;
91.77*1)8; Vˆ2770.6*6) A ; Zˆ4; Dcalcˆ1.202 g cm23
m*CuKa)ˆ17.62 cm21; total of 8238 re¯ections within
2uˆ120.18 and I.3s*I). The ®nal R factor was 4.7%
*Rwˆ5.4%).
Crystal data for 10 are as follows: C33H42Si; space group

T. Nishinaga et al. / Tetrahedron 57 '2001) 3645±3656
3655
Ê
Ê
Ê
Zang, S. Science 1994, 263, 984±985. *d) Reed, C. A.; Xie,
Z. Science 1994, 263, 985±986.
P21/n; aˆ11.9511*8) A, bˆ16.2109*9) A, cˆ14.168*1) A,
bˆ91.262*1)8; Vˆ2744.2*3) A ; Zˆ4; Dcalcˆ1.130 g cm23
;
Ê 3
m*CuKa)ˆ8.70 cm21; total of 5276 re¯ections within
9. Schleyer, P. v. R.; Buzek, P.; MuÈller, T.; Apeloig, Y.; Siehl,
H.-U. Angew. Chem., Int. Ed. Engl. 1993, 32, 1471±1473.
10. Olah, G. A.; Rasul, G.; Buchholz, H. A.; Li, X.-y.; Prakash,
G. K. S. Bull. Soc. Chim. Fr. 1995, 132, 569±574.
11. *a) Olsson, L.; Cremer, D. Chem. Phys. Lett. 1993, 215, 433±
443. *b) Olsson, L.; Ottosson, C.-H.; Cremer, D. J. Am. Chem.
Soc. 1995, 114, 7460±7479. *c) Arshadi, M.; Johnels, D.;
Edlund, U.; Ottosson, C.-H.; Cremer, D. J. Am. Chem. Soc.
1996, 118, 5120±5131.
I.2s*I). The ®nal R factor was 5.1% *Rwˆ5.0%).
Crystal data for 20 are as follows: C24H34; space group
Ê
Ê
Ê
P21/n; aˆ6.2598*9) A, bˆ18.285*3) A, cˆ15.499*2) A,
bˆ93.963*3)8; Vˆ1769.8*4) A ; Zˆ4; Dcalcˆ1.210 g cm23
;
Ê 3
m*MoKa)ˆ0.067 mm21; total of 2507 re¯ections within
I.2s*I). The ®nal R factor was 8.0% *wR2ˆ21.0%).
All of these structures were solved by direct methods and
expanded using Fourier techniques. The non-hydrogen
atoms were re®ned anisotropically. Hydrogen atoms were
geometrically calculated and ®xed.
12. *a) Lambert, J. B.; Zhao, Y. Angew. Chem., Int. Ed. Engl.
1997, 36, 400±401. *b) Lambert, J. B.; Zhao, Y.; Wu, H.;
Tse, W. C.; Kuhlmann, B. J. Am. Chem. Soc. 1999, 121,
5001±5008.
13. Steinberger, H.-U.; MuÈller, T.; Auner, N.; Maerker, C.;
Schleyer, P. v. R. Angew. Chem., Int. Ed. Engl. 1997, 36,
626±628.
Acknowledgements
14. Sekiguchi, A.; Matsuno, T.; Ichinohe, M. J. Am. Chem. Soc.
2000, 122, 11250±11251.
This work was supported by a Grant-in-Aid for Scienti®c
Research on Priority Areas *A) *No. 10146102) from the
Ministry of Education, Science, Sports and Culture, Japan.
15. Apeloig, Y.; Karni, M. In The Chemistry of Organic Silicon
Compounds; Rappoport, Z., Apeloig, Y., Eds.; Wiley:
Chichester, 1998; Vol. 2, pp. 1±101.
16. Wakita, K.; Tokitoh, N.; Okazaki, R.; Takagi, N.; Nagase, S.
J. Am. Chem. Soc. 2000, 122, 5648±5649.
References
17. *a) Tokitoh, N.; Wakita, K.; Okazaki, R.; Nagase, S.; Schleyer,
P. v. R.; Jiao, H. J. Am. Chem. Soc. 1997, 119, 6951±6952.
*b) Wakita, K.; Tokitoh, N.; Okazaki, R.; Nagase, S.;
Schleyer, P. v. R.; Jiao, H. J. Am. Chem. Soc. 1999, 121,
11,336±11,344.
1. *a) Lambert, J. B.; Kania, L.; Zhang, S. Chem. Rev. 1995, 95,
1191±1201. *b) Schleyer, P. v. R. Science 1997, 275, 39±40.
*c) Belzner, J. Angew. Chem., Int. Ed. Engl. 1997, 36, 1277±
1280. *d) Reed, C. A. Acc. Chem. Res. 1998, 31, 325±332.
*e) Maerker, C.; Schleyer, P. v. R. In The Chemistry of
Organic Silicon Compounds; Rappoport, Z., Apeloig, Y.,
Eds.; Wiley: Chichester, 1998; Vol. 2, pp. 513±555.
*f) Lickiss, P. D. In The Chemistry of Organic Silicon
Compounds; Rappoport, Z., Apeloig, Y., Eds.; Wiley:
Chichester, 1998; Vol. 2, pp. 557±594.
18. West, R.; Sohn, H.; Bankwitz, U.; Calabrese, J.; Apeloig, Y.;
Mueller, T. J. Am. Chem. Soc. 1995, 117, 11608±11609.
19. Freeman, W. P.; Tilley, D.; Yap, G. P. A.; Rheingold, A. L.
Angew. Chem., Int. Ed. Engl. 1996, 35, 882±884.
20. Gordon, M. S.; Boudjouk, P.; Anwari, F. J. Am. Chem. Soc.
1983, 105, 4972±4976.
2. *a) Lambert, J. B.; Schulz, W. J.; McConnell, J. A.; Schilf, W.
J. Am. Chem. Soc. 1988, 110, 2201±2210. *b) Lambert, J. B.;
Schilf, W. J. Am. Chem. Soc. 1988, 110, 6364±6367.
*c) Lambert, J. B.; Kania, L.; Schilf, W.; McConnell, J. A.
Organometallics 1991, 10, 2578±2584.
21. *a) Goldfuss, B.; Schleyer, P. v. R. Organometallics 1995, 14,
1553±1555. *b) Schleyer, P. v. R.; Maerker, C.; Dransfeld, A.;
Jiao, H.; Hommes, N. J. R. v. E. J. Am. Chem. Soc. 1996,
118, 6317±6318. *c) Goldfuss, B.; Schleyer, P. v. R.
Organometallics 1997, 16, 1543±1552.
3. *a) Lambert, J. B.; Zang, S.; Stern, C. L.; Huffman, J. C.
Science 1993, 260, 1917±1918. *b) Lambert, J. B.;
Zhang, S. J. Chem. Soc., Chem. Commun. 1993, 383±384.
*c) Lambert, J. B.; Zang, S.; Ciro, S. M. Organometallics
1994, 13, 2430±2443.
22. *a) Murthy, S.; Nagano, Y.; Beauchamp, J. L. J. Am. Chem.
Soc. 1992, 114, 3573±3574. *b) Nagano, Y.; Murthy, S.;
Beauchamp, J. L. J. Am. Chem. Soc. 1993, 115, 10805±10811.
23. Jarek, R. L.; Shin, S. K. J. Am. Chem. Soc. 1997, 119, 6376±
6383.
4. *a) Reed, C. A.; Xie, Z.; Bau, R.; Benesi, A. Science 1993,
262, 402±404. *b) Xie, Z.; Liston, D. J.; Jelink, T.; Mitro, V.;
Bau, R.; Reed, C. A. J. Chem. Soc., Chem. Commun. 1993,
384±386. *c) Xie, Z.; Bau, R.; Benesi, A.; Reed, C. A.
Organometallics 1995, 14, 3933±3941. *d) Xie, Z.; Manning,
J.; Reed, R. W.; Mathur, R.; Boyd, P. D. W.; Benesi, A.; Reed,
C. A. J. Am. Chem. Soc. 1996, 118, 2922±2928.
24. *a) Nicolaides, A.; Radom, L. J. Am. Chem. Soc. 1994, 116,
9769±9770. *b) Nicolaides, A.; Radom, L. J. Am. Chem. Soc.
1996, 118, 10561±10570. *c) Nicolaides, A.; Radom, L.
J. Am. Chem. Soc. 1997, 119, 11933±11937.
25. Berti, G. J. Org. Chem. 1957, 22, 230±230.
26. *a) Komatsu, K.; Akamatsu, H.; Jinbu, Y.; Okamoto, K. J. Am.
Chem. Soc. 1988, 110, 633±634. *b) Komatsu, K.; Akamatsu,
H.; Aonuma, S.; Jinbu, Y.; Maekawa, N.; Takeuchi, K.
Tetrahedron 1991, 47, 6951±6996. *c) Komatsu, K.;
Nishinaga, T.; Maekawa, N.; Kagayama, A.; Takeuchi, K.
J. Org. Chem. 1994, 59, 7316±7321.
5. *a) Prakash, G. K. S.; Kenyaniyan, S.; Aniszfeld, L.; Heiliger,
L.; Olah, G. A.; Stevens, R. C.; Choi, H. K.; Bau, R. J. Am.
Chem. Soc. 1987, 109, 5123±5126. *b) Olah, G. A.; Rasul, G.;
Heiliger, L.; Bausch, J.; Prakash, G. K. S. J. Am. Chem. Soc.
1992, 114, 7737±7742.
27. A preliminary communication has appeared: Nishinaga, T.;
Izukawa, Y.; Komatsu, K. J. Am. Chem. Soc. 2000, 122,
9312±9313.
6. Lickiss, P. D. J. Chem. Soc., Dalton Trans. 1992, 1333±1338.
7. Kira, M.; Hino, T.; Sakurai, H. Chem. Lett. 1993, 153±156.
8. *a) Pauling, L. Science 1994, 263, 983±983. *b) Olah, G. A.;
Rasul, G.; Li, X.; Buchholz, H. A.; Sandford, G.; Prakash,
G. K. S. Science 1994, 263, 983±984. *c) Lambert, J. B.;
28. Nishinaga, T.; Izukawa, Y.; Komatsu, K. Chem. Lett. 1998,
269±270.

3656
T. Nishinaga et al. / Tetrahedron 57 '2001) 3645±3656
29. Komatsu, K.; Aonuma, S.; Jinbu, Y.; Tsuji, R.; Hirosawa, C.;
Takeuchi, K. J. Org. Chem. 1991, 56, 195±203.
32. *a) Nishinaga, T.; Komatsu, K.; Sugita, N. J. Chem. Soc.,
Chem. Commun. 1994, 2319±2320. *b) Nishinaga, T.;
Kawamura, T.; Komatsu, K. J. Org. Chem. 1997, 62, 5354±
5362. *c) Nishinaga, T.; Wakamiya, A.; Komatsu, K. Chem.
Commun. 1999, 777±778.
30. gaussian 98, Revision A.5, Frisch, M. J.; Trucks, G. W.;
Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman,
J. R.; Zakrzewski, V. G.; Montgomery, Jr., J. A.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.;
Tomasi, J.; Barone,V.; Cossi, M.; Cammi, R.; Mennucci, B.;
Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson,
G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.;
Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill,
P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.;
Gonzalez, C.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A.
Gaussian, Inc., Pittsburgh PA, 1998.
33. For the method of calculations see: Subramanian, G.;
Schleyer, P. v. R.; Jiao, H. Organometallics 1997, 16,
2362±2369.
34. Kira, M.; Hino, T.; Sakurai, H. J. Am. Chem. Soc. 1992, 114,
6697±6700.
35. Johannsen, M.; Jùrgensen, K. A.; Helmchen, G. J. Am. Chem.
Soc. 1998, 120, 7637±7638.
36. Olah, G. A.; Rasul, G.; Prakash, G. K. S. J. Am. Chem. Soc.
1999, 121, 9615±9617.
37. Bahr, S. R.; Boudjouk, P. J. Am. Chem. Soc. 1993, 115, 4514±
4519.
38. Chien, J. C. W.; Tsai, W.-M.; Rausch, M. D. J. Am. Chem.
Soc. 1991, 113, 8570±8571.
31. MuÈller, T.; Zhao, Y.; Lambert, J. B. Organometallics 1998,
17, 278±280.