1,2- versus 1,3-Silyl migration in the reaction of acylsilanes with silyl-substituted carbanions

Article abstract of DOI:10.1016/j.tetlet.2008.02.077

The Peterson olefination of acylsilanes 1 by silylated carbanions 5,14 to give alkenes 13,17 via a 1,3-silyl shift is favored by the stabilization of the carbanion intermediate 16 through two sulfide units versus one as in 12, while a combination of DMPS on 1 and TBS on 14 leads to 1,2-migration product 15.

Full text of DOI:10.1016/j.tetlet.2008.02.077

Available online at www.sciencedirect.com
Tetrahedron Letters 49 (2008) 2628–2630
1,2- versus 1,3-Silyl migration in the reaction of acylsilanes
with silyl-substituted carbanions
Thomas Puschke, Jens Lange, Ernst Schaumann *
Institute of Organic Chemistry, Clausthal University of Technology, Leibnizstrasse 6, 38678 Clausthal-Zellerfeld, Germany
Received 20 December 2007; revised 13 February 2008; accepted 15 February 2008
Available online 19 February 2008
Abstract
The Peterson olefination of acylsilanes 1 by silylated carbanions 5,14 to give alkenes 13,17 via a 1,3-silyl shift is favored by the
stabilization of the carbanion intermediate 16 through two sulfide units versus one as in 12, while a combination of DMPS on 1 and
TBS on 14 leads to 1,2-migration product 15.
Ó 2008 Published by Elsevier Ltd.
Keywords: Silyl shift; Acylsilane; Carbanion; Peterson olefination
Nu
C
Nu
C
1,2-C ? O Silyl migrations have emerged as a powerful
tool to generate a reactive carbanion intermediate 3 by
nucleophilic attack on acylsilane 1 via the initial adduct 2
(Scheme 1, Eq. a).1,2 Similarly, a 1,3-C ? O silyl migration
is the crucial step in the base-induced Peterson olefination
of a carbonyl compound 4 to give alkene 7 (Scheme 1, Eq.
b).3,4 For the latter pathway, an alternative mechanism
involves (2+2) cycloreversion at the stage of an intermedi-
ate 1,2-oxasiletanide to give the alkene product.3,5 How-
ever, in the presence of electron-withdrawing substituents
such as the sulfide units of the present study, the evidence
suggests that the basic elimination pathway is not con-
certed, but includes an intermediate of type 6.3
R1
C
SiR23
+
(a)
products
R1
SiR23
R1
Nu
O
O
OSiR3
1
2
3
R3
R4
R4
R5
R1
C
R3
+
R1
R5
C
SiR63
O
O
SiR63
4
5
R3
R4
R4
R3
~SiR663
R1
R5
(b)
-
OSiR63
R5
R1
OSiR36
An aspect that has been widely neglected is the potential
competition between a 1,2-silyl shift and a 1,3-silyl shift in
the reaction of acylsilane 1 with a silicon-substituted carb-
anion 5 as nucleophile Nu (Scheme 2). The only available
evidence is the slight preference for the 1,3 shift (Peterson
olefination) in the reaction of 1-TMS with lithiated silyl-
thioanisole 5-TMS to give vinylsilane 13-TMS along with
acetophenone apparently as the secondary product of a-
6
7
Scheme 1.
silyl ketone 11 (Table 1, entry 1).6 In our hands, this reac-
tion is not clean; after extensive chromatography (SiO2,
ether/hexane) we did find silylketone 11,7 but also aceto-
phenone as secondary product (combined yield, 27%);
Peterson product 13-TMS was obtained in the same yield.
A handle to influence the ratio of 1,2- versus 1,3-silyl
shifts may be the use of different residues R2, R6 (Scheme
1) on the silicon resulting in different migratory aptitudes.
Abbreviations: TBS, SiMe2tBu; DMPS, SiMe2Ph
*
Corresponding author. Tel.: +49 5323 722519; fax: +49 5323 722858.
E-mail address: Ernst.schaumann@tu-clausthal.de (E. Schaumann).
0040-4039/$ - see front matter Ó 2008 Published by Elsevier Ltd.
doi:10.1016/j.tetlet.2008.02.077

T. Puschke et al. / Tetrahedron Letters 49 (2008) 2628–2630
2629
Ph
C
SiR23
At the same time, the yield of the Peterson product 13 goes
down to ca. 13% (Table 1).
O
(R1 = Ph)
The ratio of 1,2- versus 1,3-silyl shift may also be influ-
enced by the stability of the intermediate carbanion 12 as
formed from 8 by 1,3-silyl migration. We have successfully
employed lithiated silyl thioacetals like 14 in the reaction
with functionalized oxiranes to give eventually cyclo-
alkanes.8,9 Using 14 should give, after the 1,3 shift, the
stabilization of the negative charge by two sulfur substitu-
ents as in 16 versus one phenylthio group in 12, but should
be without appreciable effect on intermediate 9 of the 1,2-
silyl migration route (Schemes 1–3). In fact, in a much clea-
ner reaction, acylsilane 1-TMS and lithiated trimethyl-
silylthioacetal 14-TMS give a high yield of alkene 17-
TMS while a ketone of type 15 or the corresponding
desilylated species could not even be detected in trace
amounts by GC–MS combination (Table 1, entry 3).11
Furthermore, we again checked the influence of the silyl
residue on the ‘regioselectivity’ of silyl migration. Here, we
assumed that a tBu group in the SiR63 unit should suppress
the 1,3 shift. However, the increased bulk in carbanion
14-TBS as compared to 14-TMS widely slows down the
addition to acylsilane 1-TMS and allows only minor
conversion. But it could be established by GS–MS analysis
that products 15, 17-TMS are formed in equal amounts
demonstrating that the 1,2 shift is encouraged relative to
the use of carbanion 14-TMS (Table 1, entry 4).
1
PhS
Ph
SiMe3
SiR23
1-TMS : R2 = Me
1-DMPS : SiR23 = Me2Ph
+
PhS
SiMe3
O
8
(R4 = PhS, R5 = H, R6 = Me)
1,2 - SiR32
5
1,3 - SiMe3
PhS
Ph
PhS
Ph
SIMe3
SiR23
OSiMe3
12
9
OSiR23
PhS
-
PhS
SiMe3
10
13
Ph
SiR23
Ph
OSiR23
work-up
H3O
-
-
-TMS : R2 = Me
9 12 13
, ,
8
8
-DMPS : R23 = Me2Ph
12 13
,
9
,
SiMe3
On the other hand, the attempt to enhance the contribu-
tion of the 1,2 shift by using acylsilane 1-DMPS with the
most easily migrating silyl group and carbanion 14-TMS
11
Ph
O
Scheme 2.
MeS
Ph
TBS
O
1,2 silyl
shift
Actually, the study of the reaction of bis-silylated thioani-
sole (5 with R4 = SPh, R5 ¼ SiR73) with oxiranes had
revealed the following sequence of increased ease of migra-
tion:8–10
for
R63 Si = TBS
MeS
MeS
1 TMS
-
15
+
C
or
1-DMPS
MeS
Ph
SMe
SiR63
MeS
Ph
SMe
SiR23
1,3 silyl
shift
SiMe2tBuðTBSÞ ꢀ SiMePh2 < SiMe3 < SiMe2PhðDMPSÞ
14 TMS
14 TBS
-
-
: R6 = Me
SiR23
: R63 = tBuMe2
So, we assumed that a Ph group in the SiR23 unit of the
acylsilane 1-DMPS should favor the 1,2 shift. In fact, the
reaction of 1-DMPS with carbanion 5-TMS gives an in-
creased yield of acetophenone, the desilylation product of
the 1,2-silyl shift pathway (Scheme 2; Table 1, entry 2).
OSiR63
17
16
17 TMS
-
: R2 = Me
17-DMPS
: R23 = Me2Ph
Scheme 3.
Table 1
Competition experiments in the reaction of acylsilanes 1 with carbanions 5, 14 (counterion Li)
Entry
Acylsilane
Carbanion
Reaction time [h] (temp)
Product 11, 15
Product 13
1
2
3
4
5
6
a
1-TMS
1-DMPS
1-TMS
1-TMS
1-DMPS
1-DMPS
5-TMS
5-TMS
14-TBS
14-TBS
14-TMS
14-TBS
0.3 (0 °C)
0.3 (0 °C)
12 (ꢁ78 °C), then 10 (20 °C)
12 (ꢁ78 °C), then 28 (20 °C)
12 (ꢁ78 °C), then 5 (20 °C)
12 (ꢁ78 °C), then 48 (20 °C)
27%a
42%b
—
15 (trace)
—
13-TMS (27%)
13-DMPS (13%)
17-TMS (84%)
17-TMS (trace)c
17-DMPS (49%)
—
15 (42%)
Product contaminated with acetophenone from silyl ketone 11.
Only acetophenone as desilylation product.
GC–MS analysis showed equal amounts of 15 and 17-TMS.
b
c

2630
T. Puschke et al. / Tetrahedron Letters 49 (2008) 2628–2630
failed as shown by the formation of alkene 17-DMPS as
the only detectable product in a complex reaction mixture
(Table 1, entry 5). Obviously, carbanion stabilization in
intermediate 16 overrules the effect of silicon migratory
aptitude.
In contrast, when additional hindrance to the 1,3 shift
by using 14-TBS is introduced, a total swing to the 1,2-silyl
shift is observed (Table 1, entry 6). Obviously, in spite of
the steric hindrance, small amounts of the primary addition
product of type 8 give rapid DMPS migration pulling the
reaction to the observed ketone product.
References and notes
1. Moser, W. H. Tetrahedron 2001, 57, 2065–2084.
2. Patrocinio, A. F.; Moran, P. J. S. J. Braz. Chem. Soc. 2001, 7–31.
3. Ager, D. J. Org. React. 1990, 38, 1–223.
4. van Staden, L. F.; Gravestock, D.; Ager, D. J. Chem. Soc. Rev. 2002,
31, 195–200.
5. Naganuma, K.; Kawashima, T.; Okazaki, R. Chem. Lett. 1999, 1139–
1140.
6. Kang, K.-T.; Sung, T. M.; Lee, K.-R.; Lee, J. G.; Jyung, K. K. Bull.
Korean Chem. Soc. 1993, 14, 757–759.
7. Furstner, A.; Kollegger, G.; Weidmann, H. J. Organomet. Chem.
¨
1991, 414, 295–305.
Thus, it could be demonstrated that the ambiguity
between 1,2- and 1,3-silyl shift in the reaction of acylsilanes
1 with carbanions 5,14 can be further supported or totally
suppressed by the adequate choice of the carbanion com-
ponent and of the residues R2, R6 on the silicon.
8. Schaumann, E.; Kirschning, A. Synlett 2007, 177–190.
9. Schaumann, E. Heteroat. Chem. 2007, 18, 600–608.
10. Bra¨uer, N.; Michel, T.; Schaumann, E. Tetrahedron 1999, 54, 11481–
11488.
11. Standard reaction conditions: THF, generation of carbanion 5 (14)
from the corresponding CH acid (1.0 equiv) with butyl lithium
(1.1 equiv) at ꢁ78 °C (0.5 h) and 0–20 °C (0.5 h). Addition of
acylsilane 1 (1 equiv) at ꢁ78 °C and stirring (see Table 1), finally
allowing the mixture to reach room temperature (overnight). If the
reaction was incomplete, stirring was continued (Table 1) and finally
the reaction mixture was quenched with water. After the usual
work-up, products 11, 13, 15, 17 were isolated by flash chromato-
graphy (silica or Al2O3, eluent petroleum ether with 0–5% ethyl
acetate).
Acknowledgments
The authors thank the Deutsche Forschungsgemein-
schaft (Scha 231/11) and the Fonds der Chemischen
Industrie, Frankfurt, for financial support. We thank
Mrs. Monika Ries for skillful experimental assistance.