Cleavage of Si-Ar bond vs Si-Me bond: A remarkable counterion effect on reactivity

Article abstract of DOI:10.1039/b205373j

Formation of distinctly different products from the same alkoxide intermediate indicates a strong dependence of reaction pathways on counterions.

Full text of DOI:10.1039/b205373j

Cleavage of Si–Ar bond vs Si–Me bond: a remarkable counterion effect
on reactivity†
Suresh Kumar Tipparaju, Sunil K. Mandal,‡ Surojit Sur,§a Vedavati G. Puranik and Amitabha
Sarkar¶
a
a
a
b
a
Division of Organic Chemistry, National Chemical Laboratory, Pune 411 008, India.
E-mail: ocas@mahendra.iacs.res.in
Division of Physical Chemistry, National Chemical Laboratory, Pune 411 008, India
b
Received (in Cambridge, UK) 6th June 2002, Accepted 16th July 2002
First published as an Advance Article on the web 30th July 2002
Formation of distinctly different products from the same
alkoxide intermediate indicates a strong dependence of
reaction pathways on counterions.
1
We have previously described that nucleophile-assisted Ar–Si
bond cleavage is particularly facile when the arene ring is
complexed with a Cr(CO)
aryl anion intermediate anchored on Cr(CO)
3
fragment (Chart 1). Stability of the
moiety is believed
3
to facilitate such desilylation. This reaction provides a very
useful method for introducing electron-withdrawing substi-
tuents on the arene ring by an ipso substitution of the SiMe
group. The facility of Si–Ar bond cleavage is evident also in a
3
2
3
recently reported observation of facile Brook rearrangement
1,4+C-to-O) of the alkoxide resulting from addition of
methyllithium to o-SiMe –benzaldehyde–Cr(CO) complex
Chart 1).
The present paper describes a dramatic departure from this
Chart 1
(
3
3
reaction was carried out at 278 °C and the reaction mixture was
(
quenched immediately after complete consumption of starting
material (15–30 min), the carbinol 2a was isolated in 39% yield
seemingly simple picture. We report that a change of the
counterion of the alkoxide intermediate (e.g. replacing lithium
with magnesium) leads to an unexpected but decisive alteration
of reaction course.
as a diastereomerically pure product6 (path a, Scheme 1).
Unexpectedly, a conjugate addition product 2A was also
obtained from the same reaction mixture in 47% yield as a pure
diastereomer. Clean conversion of 2a to 3 in presence of BuLi
In continuation of our on-going investigations with arene–
(
path c, Scheme 1) established that the conjugate addition
4
tricarbonylchromium complexes, we carried out addition of
product 2A was not formed by an anionic oxy-Cope rearrange-
ment of the intermediate alkoxide during allyllithium addition
to 1; rather, it resulted from a competitive parallel reaction.
When the reaction was allowed to warm up to room temperature
after the addition of allyllithium and stirred for several hours,
complex 3 resulting from a Brook rearrangement was obtained
instead of carbinol 2a in 40% yield (path b, Scheme 1) along
with some 2A (45%).
allyllithium to an acyclic enone 1 (Scheme 1).5 When the
†
Electronic supplementary information (ESI) available: spectroscopic data
and experimental section. See http://www.rsc.org/suppdata/cc/b2/
b205373j/
Present address: University of Utah, Salt Lake City, USA.
Present address: John Hopkins University, Baltimore, USA.
Present address: Indian Association for the Cultivation of Science,
‡
§
¶
On the other hand, reaction of allylmagnesium bromide and
complex 1 at 278 °C for 20 min showed poor diaster-
Kolkata, India.
Scheme 1
This journal is © The Royal Society of Chemistry 2002
1
924
CHEM. COMMUN., 2002, 1924–1925

eoselectivity and afforded isomeric carbinols 2a and 2b (as 1+1
mixture) in excellent yield (96%). When the reaction mixture of
allylmagnesium bromide and complex 1 was allowed to slowly
attain room temperature over several hours, isomeric five-
membered heterocycles (4a and 4b) were the only isolable
lead to two mutually exclusive products seemingly as a result of
kinetic selection. These results once again raise important
questions as to the exact role of counterions—the nature of ion-
pair, solvation, structure and stability of pentacoordinated
silicon intermediates, factors that precisely tune the energetics
of different mechanistic possibilities. Our current endeavors
strive to address some of these subtle yet significant issues, and
explore the scope and generality of these observations to invent
useful synthetic applications.
The authors (S. K. T., S. K. M. and S. S.) are grateful to CSIR,
New Delhi, for award of research fellowships, and Reliance
Industries Ltd. for financial support. We thank Professor T. K.
Sarkar, IIT, Kharagpur, India, for valuable suggestions.
7
products (path d, Scheme 1). Two distinct three-proton singlets
corresponding to the two Si–Me groups at 0.50 and 0.66 ppm
are the diagnostic proton NMR features for isomer 4a. For 4b
these peaks appear at 0.45 and 0.68 ppm. Fortuitously, they
could be readily separated by fractional crystallization and the
structure of isomer 4a was confirmed by crystal structure
8
determination (Fig. 1). No Brook product was obtained in this
reaction.
Notes and references
1
(a) S. K. Mandal and A. Sarkar, J. Org. Chem., 1998, 63, 1901; (b) S.
K. Mandal and A. Sarkar, J. Org. Chem., 1998, 63, 5672.
2
(a) W. H. Moser, K. E. Endsley and J. T. Colyer, Org. Lett., 2000, 2,
7
17; (b) W. H. Moser, J. Zhang, C. S. Lecher, T. L. Frazier and M. Pink,
Org. Lett., 2002, 4, 1981.
3
4
Review: W. H. Moser, Tetrahedron, 2001, 57, 2065.
For an account, see: A. Sarkar, S. Ganesh, S. Sur, S. K. Mandal, V. M.
Swamy, B. C. Maity and T. Sureshkumar, J. Organomet. Chem., 2001,
6
24, 18.
5
6
All compounds reported herein are racemic complexes, only one
enantiomer is depicted in the schemes to illustrate the chemistry. Yields
refer to isolated yields of purified products.
The stereochemical assignments in the present paper are based on close
similarity of spectral data among the analogs of 2a and 2A and crystal
structure of an analog of 2a that have been recently reported: S. Sur, S.
K. Mandal, S. Ganesh, V. G. Puranik and A. Sarkar, Indian J. Chem.,
Sect. B, 2001, 40, 1063.
Fig. 1 ORTEP diagram of 4a. Atoms are drawn at 50% probability.
That the heterocyclic product 4 was indeed formed by a
pathway dictated by magnesium counterion was readily ascer-
tained by treatment of lithium alkoxide A with a three-fold
7 For examples of cyclization by reaction of oxygen nucleophile on a
trimethylsilyl group, see: (a) Y. M. Hijji, P. F. Hudrlik and A. M.
Hudrlik, Chem Commun., 1998, 1213; (b) P. F. Hudrlik, Y. M. Abdallah
and A. M. Hudrlik, Tetrahedron Lett., 1992, 33, 6747; (c) Y.
Yamamoto, Y. Takeda and K. Akiba, Tetrahedron Lett., 1989, 30, 725;
2 2
excess of MgBr –Et O prior to warming up, which yielded 4a
instead of 2a or 3 as the product (path e, Scheme 1). Also,
addition of one equivalent of benzophenone in the reaction
mixture after allylmagnesium bromide had completely con-
sumed substrate 1 but before warming up, afforded diphenyl
methyl carbinol in 74% yield (yield of 4 was 82%), thereby
accounting for the methyl group that departed from silicon as a
nucleophile.
Based on the above observations, we propose that the
reaction proceeds via the common alkoxide intermediate A,
which can exist in equilibrium with a cyclic intermediate B
featuring a pentacoordinated silicon9 (Scheme 1). At low
temperature the equilibrium favors A. When lithium is the
counterion, and the temperature is raised, Ar–Si bond cleaves to
produce an aryl anionic species stabilized by tricarbonylchro-
mium complexation, with concomitant formation of O–Si bond,
affording the Brook product. It could occur in a stepwise
(d) W. Kirmse and F. Soellenboehmer, J. Chem. Soc., Chem. Commun.,
1
989, 774; (e) K. Tamao, T. Nakajima, R. Sumiya, H. Arai, N. Higuchi
and Y. Ito, J. Am. Chem. Soc., 1986, 108, 6090; (f) M. Uemura, T.
Kobayashi, K. Isobe, T. Minami and Y. Hayashi, J. Org. Chem., 1986,
51, 2859; (g) H. Oda, M. Sato, Y. Morizawa, K. Oshima and H. Nozaki,
Tetrahedron., 1985, 41(16), 3257.
(a) Needle like yellow single crystals were grown in a mixture of
dichloromethane and pet ether. Data were collected on a MACH-3
diffractometer using Mo-Ka radiation with fine focus tube.
8
C
23
H
22CrO
group P2 /n, a = 12.488(2), b = 9.1030(9), c = 20.682(3) Å, b =
07.16(1)°, V = 2246.4(5) Å , Z = 4, D
4
Si, M = 442.50. Crystals belong to monoclinic, space
1
3
23
1
c
= 1.308 Mg m , m(Mo-Ka)
0.587 mm , T = 293(2) K, 3546 unique [I > 2s(I)], R = 0.0341,
wR2 = 0.0811. All the data were corrected for Lorentzian, polarisation
21
=
1
8b
and absorption effects. SHELX-97 (SHELXTL) was used for
structure solution and full matrix least squares refinement on F . CCDC
2
manner or by a concerted, intramolecular, S
N
i type displace-
ment10 at silicon by the alkoxide. The driving force behind
187534. See http://www.rsc.org/suppdata/cc/b2/b205373j/ for crys-
tallographic files in .cif or other electonic format (b) G. M. Sheldrick,
SHELX-97 program for crystal structure solution and refinement,
University of Goettingen, Germany, 1997.
For an extensive collection of references including reviews on
hypervalent silicon and other elements, see: T. Kawashima, K.
Naganuma and R. Okazaki, Organometallics, 1998, 17, 367.
preferential rupture of a Si–Me bond mediated by magnesium
counterion, on the other hand, appears much less obvious at this
point.11 The intermediate B seems significant only when the
counterion is magnesium and temperature is raised from 278
9
°
C. Formation of intermediate B is certainly facilitated12 by
gem-disubstitution at the chiral center (cf. Thorpe-Ingold
10 E. W. Colvin, in Silicon in Organic Synthesis, Butterworths, 1981, p.
42.
11 We have found that tricarbonylchromium plays no role in this
cyclization. Uncomplexed substrates also similarly cyclize in the
presence of magnesium counterion (Suresh Kumar Tipparaju, un-
published results).
effect). One is prompted to invoke an analogy from the
literature13 to suggest a four-centered transition state for facile
expulsion of MeMgBr to complete—simultaneously—the het-
erocycle formation. Addition of methyllithium to 4 reverses the
steps of heterocyclization to provide the lithium alkoxide A,
eventually preferring Brook rearrangement as the kinetically
favored pathway at warmer temperatures.
In summary, we presented here an interesting example of
counterion dependence of reaction pathways involving argu-
ably the same intermediate alkoxide. Two different counterions
1
2 A gem-dimethyl substitution is equally effective. The lithium alkoxide
derived from (CO) Cr-[o-TMS-C -C(Me) OH] leads to Brook
3
H
6 4
2
product, while magnesium alkoxide promotes cyclization (Suresh
Kumar Tipparaju, unpublished results).
13 J. Wong, K. A. Sannes, C. E. Johnson and J. I. Brauman, J. Am. Chem.
Soc, 2000, 122, 10878.
CHEM. COMMUN., 2002, 1924–1925
1925