Mechanistic studies of the allylic rearrangements of α-silyloxy allylic silanes to silyloxy vinylic silanes

Article abstract of DOI:10.1021/jo070634p

(Chemical Equation Presented) Mechanistic evidence suggests that the Lewis acid-promoted allylic rearrangement of α-silyloxy allylic silanes proceeds along an ionic reaction pathway involving a contact ion pair. The driving force for this transformation i

Full text of DOI:10.1021/jo070634p

unexpected considering that the R-acetoxy allylic silane could
rearrange by a concerted [3,3] sigmatropic process or through
an ionization mechanism that proceeds through a less basic
acetate counterion. To discover the origins of the increased
reactivity of the R-silyloxy allylic silane, a detailed mechanistic
study was undertaken.
A crossover experiment between two distinct R-silyloxy
allylic silanes demonstrated that the rearrangement was an
intramolecular process. Treating a mixture of allylic silanes 1
and 3 with BF3‚OEt2 afforded only vinylic silanes 2 and 4 (eq
2); crossover products 5 and 6 were not observed. The crossover
Mechanistic Studies of the Allylic
Rearrangements of r-Silyloxy Allylic Silanes to
Silyloxy Vinylic Silanes
Angie I. Kim, Kyle L. Kimmel, Antonio Romero,
Jacqueline H. Smitrovich, and K. A. Woerpel*
Department of Chemistry, UniVersity of California,
IrVine, California 92696-2025
kwoerpel@uci.edu
ReceiVed March 27, 2007
experiment was repeated in a variety of polar solvents, including
2-nitropropane, and at different concentrations (0.014 and 1.4
M) in CH2Cl2.5 In all cases, increased solvation did not provide
the crossover products, verifying that the transformation was
an intramolecular rearrangement.
An asymmetric variant of the R-silyloxy allylic silane
rearrangement established that the rearrangement proceeds via
a contact ion pair. Treatment of enantioenriched allylic silane
(S)-1 with BF3‚OEt2 afforded the vinylic silane (R)-2 with a
diminished enantiomeric excess (eq 3).6 The absolute stereo-
chemistry of the major enantiomer of the vinylic silane was
deduced from the known deprotected alcohol.7 The partial
retention observed, coupled with a lack of crossover products,
suggested an ionic mechanism was operating through a contact
ion pair.
Mechanistic evidence suggests that the Lewis acid-promoted
allylic rearrangement of R-silyloxy allylic silanes proceeds
along an ionic reaction pathway involving a contact ion pair.
The driving force for this transformation is alleviation of
steric congestion at the allylic position of the R-silyloxy
allylic silane and stabilization of πcc by hyperconjugation.
Although R-oxygenated allylic silanes are useful reagents for
organic synthesis,1,2 their utility has been limited because of
their sensitivity toward decomposition. It has been reported that
both R-acetoxy allylic silanes and R-hydroxy allylic silanes
undergo rearrangements under acidic conditions.3,4 In this Note,
we demonstrate that R-silyloxy allylic silanes undergo rapid
allylic rearrangements to provide R-silyloxy vinylic silanes.
These reactions are faster than the rearrangements of other
R-oxygenated allylic silanes. Mechanistic studies suggest that
ionization to form a contact ion pair is rate-determining, and
the rate of ionization depends upon steric effects more than it
depends upon electronic effects.
A study of R-silyloxy allylic silanes as nucleophilic partners
in [3+2] annulations1 led to the discovery of a rearrangement
involving the R-silyloxy allylic silanes. Treating R-silyloxy
allylic silane 1 with BF3‚OEt2 resulted in the formation of
silyloxy vinylic silane 2 in 84% yield (eq 1). A comparable
Control experiments provided additional support for the
contact ion pair mechanism. To discount racemization occurring
prior to the rearrangement, the stereochemistry scrambling
experiment was repeated and stopped prior to completion (eq
4). Analysis of the recovered starting material revealed no loss
of stereochemical information,8 indicating that racemization was
(5) Rearrangements were carried out successfully in CH2Cl2 and 2-ni-
tropropane. Other solvents such as toluene and propionitrile provided the
rearrangement, but other products were also formed.
(6) Analysis of the isolated vinylic silane indicates that the partial chirality
transfer occurred with retention of configuration. The analysis consisted of
deprotection of the vinylic silane to afford the alcohol S6 and comparing it
to the known7 alcohol (R)-S6, using chiral HPLC analysis.7
rearrangement involving R-acetoxy allylic silanes was previously
reported with use of BF3‚OEt2 catalysis at higher temperatures.3
A direct comparison between the two substrates showed that
the R-silyloxy allylic silanes were more reactive than R-acetoxy
allylic silanes toward this rearrangement. This discovery was
(1) Romero, A.; Woerpel, K. A. Org. Lett. 2006, 8, 2127-2130.
(2) Perrone, S.; Knochel, P. Org. Lett. 2007, 9, 1041-1044.
(3) Panek, J. S.; Sparks, M. A. J. Org. Chem. 1990, 55, 5564-5566.
(4) Sakaguchi, K.; Higashino, M.; Ohfune, Y. Tetrahedron 2003, 59,
6647-6658.
(7) Smitrovich, J. H.; Woerpel, K. A. J. Org. Chem. 2000, 65, 1601-
1614.
(8) Analysis was performed with the use of optical rotations because
the allylic silanes could not be resolved by HPLC. Attempts to deprotect
allylic silanes were unsuccessful due to a competing Brook rearrangement.
10.1021/jo070634p CCC: $37.00 © 2007 American Chemical Society
Published on Web 07/24/2007
J. Org. Chem. 2007, 72, 6595-6598
6595

SCHEME 1. Proposed Mechanism for Allylic
Rearrangement
Steric effects were found to play a significant role in the rate
of the reaction. Previous studies had demonstrated that R-ac-
etoxy allylic silanes rearrange at 24 °C under BF3‚OEt2 catalysis
to afford R-acetoxy vinylic silanes.3 Exposing R-acetoxy allylic
silane 7a to the rearrangement conditions yielded no rearrange-
ment (eq 6). Treatment of the more sterically encumbered
R-pivalyloxy allylic silane 7b with BF3‚OEt2 at -78 °C for 3
h afforded pivalyloxy vinylic silane 8b in 20% yield (eq 7).
These experiments support the hypothesis that steric interactions
activate the R-silyloxy allylic silanes toward rearrangement.13
not occurring before the rearrangement. To discount the
possibility of racemization occurring after the rearrangement,
enantiomerically enriched vinylic silane (S)-2 was subjected to
the rearrangement conditions for 3 h (eq 5). The recovered
vinylic silane revealed no loss of stereochemical information,
so racemization does not occur after the rearrangement.
Steric and electronic preferences drive the Lewis acid-
catalyzed rearrangement to the thermodynamically preferred
R-silyloxy vinylic silane. Ab initio calculations illustrate a
thermodynamic preference for recombination of the contact ion
pair at the γ-position over the R-position.14 This preference
arises from alleviation of steric congestion.15 Ab initio calcula-
tions also demonstrate a thermodynamic preference of vinylic
silanes over allylic silanes.16 These calculations are in good
agreement with previous reports that attribute the stabilization
of vinylic silanes to hyperconjugation of the πCCfσ*Si-C
orbitals.17,18 The experimental data and ab initio calculations
show that the driving force for this rearrangement is relief of
steric interactions and hyperconjugation.
Both the crossover and the stereochemistry scrambling
experiments suggest an intramolecular ionic mechanism con-
sistent with a contact ion pair. According to these data, we
propose the mechanism illustrated in Scheme 1. Complexation
of the Lewis acid9 would afford intermediate A. This intermedi-
ate can adopt a conformation that minimizes A1,3 strain10 and
allows for πCCfσ*CO donation likely required for ionization.11
Dissociation of the silyl ether from A affords a Lewis acid-
complexed silanol and an allylic cation. Accessing the allylic
cation through the lowest energy conformer A provides the (E)-
allylic cation, which would lead to the (E)-olefin.12 Recombina-
tion then affords the thermodynamically preferred vinylic silane
(vide infra) with retention because the Lewis acid-complexed
silanol recombines on the same face of the cation on which it
is formed. This mechanism agrees with our experimental data
and may also be used to explain the R-acetoxy rearrangement
reported previously.3
(13) Attempts to influence the rearrangement by varying the size of the
silyl substituent were unsuccessful. Utilizing the Me3Si-protected allylic
silane led to the deprotected R-hydroxy allylic silane. Utilizing the i-Pr3-
Si-protected allylic silane, the rearrangement proceeded at the same rate,
within experimental error, as the t-BuMe2Si-protected allylic silane.
(14) The energetic difference between 1-trimethylsilyl-1-trimethylsilyl-
oxybutane and 1-trimethylsilyl-3-trimethylsilyloxybutane was calculated
using ab initio calculations performed at the HF/6-31G* level. 1-Trimeth-
ylsilyl-3-trimethylsilyloxybutane was lower in energy, presumably for steric
reasons.
(15) To ensure the energy difference between 1-trimethylsilyl-1-tri-
methylsilyloxybutane and 1-trimethylsilyl-3-trimethylsilyloxybutane was due
to steric and not electronic effects, ab initio calculations were performed
for 2,2-dimethyl-3-trimethylsilyloxyhexane and 2,2-dimethyl-5-trimethyl-
silyloxyhexane at the HF/6-31G* level. The calculation showed the same
preference for both the tert-butyl and Me3Si groups, suggesting that this
was indeed a steric phenomenon and not an electronic one.
(9) For evidence supporting Lewis acid chelation of an OSiMe2t-Bu group
see: Chen, X.; Hortelano, E. R.; Eliel, E. L.; Frye, S. V. J. Am. Chem.
Soc. 1992, 114, 1778-1784.
(10) Possible conformations C1-C3 were calculated by using ab initio
calculations performed at the HF/6-31G* level, using Macspartan (Wave-
function, Irvine, CA). Conformation C1 was the lowest energy conformer
by 0.51 kcal/mol relative to C2. Conformer C3 was found to not be an
energy minimum.
(16) To determine if any stereoelectronic preference existed for the
alkene, calculations were carried out between but-2-enyltrimethylsilane and
but-1-enyltrimethylsilane. A preference was found for vinylic silanes over
allylsilanes presumably due to hyperconjugation in the vinylic silane.18 Ab
initio calculations were performed at the HF/6-31G* level.
(11) The orbital overlap of πCCfσ*CO is much better in C1 than C2
because the two orbitals are orthogonal to each other in C2.
(12) The barrier to rotation of the allyl cation is estimated to be 36-38
kcal/mol: (a) Wiberg, K. B.; Breneman, C. M.; LePage, T. J. J. Am. Chem.
Soc. 1990, 112, 61-72. (b) Gobbi, A.; Frenking, G. J. Am. Chem. Soc.
1994, 116, 9275-9286.
(17) (a) Bock, H.; Seidl, H. J. Am. Chem. Soc. 1968, 90, 5694-5700.
(b) Molle´re, P.; Bock, H.; Becker, G.; Fritz, G. J. Organomet. Chem. 1972,
46, 89-96. (c) Horn, M.; Murrell, J. N. J. Organomet. Chem. 1974, 70,
51-60.
(18) (a) Giordan, J. C.; Moore, J. H. J. Am. Chem. Soc. 1983, 105, 6541-
6544. (b) Giordan, J. C. J. Am. Chem. Soc. 1983, 105, 6544-6546.
6596 J. Org. Chem., Vol. 72, No. 17, 2007

2.5 M solution in hexanes)]. The mixture was stirred for 30 min at
-80 °C then diluted with acetic acid (0.11 g, 1.9 mmol) in THF (2
mL). The resultant solution was poured into half-saturated aqueous
NH4Cl (50 mL) and extracted with Et2O (3 × 50 mL). The
combined organic phases were washed with saturated aqueous NaCl
(75 mL), dried over Na2SO4, and concentrated in vacuo to afford
a bright yellow oil. Purification by flash chromatography (97:3
hexanes/EtOAc to 80:20 hexanes/EtOAc) afforded R-hydroxy allylic
silane (0.325 g, 54%) as a slightly yellow oil. The R-hydroxy allylic
silane was isolated in g99% ee by chiral HPLC (Chiralcel OD-H
column, 99:1 hexanes/IPA, 1 mL/min, 254 nm). The oil was
dissolved in DMF (1.0 mL) and TBDMSCl (0.250 g, 1.66 mmol)
and imidazole (0.221 g, 3.24 mmol) were added. The reaction
mixture was stirred for 12 h, at which point hexanes (10 mL) was
added and the resultant solution was washed with saturated aqueous
NH4Cl (10 mL). The layers were separated and the aqueous layer
was re-extracted with hexanes (3 × 8 mL). The combined organic
layers were washed with saturated aqueous NaCl (10 mL). The
resultant organic phase was dried over MgSO4 and filtered. The
filtrate was concentrated in vacuo to afford the R-silyloxy allylic
silane as a light yellow oil. Purification by flash chromatography
(hexanes) afforded (-)-1 as a colorless oil (0.080 g, 16%). The
product was identical with (()-1 by 1H NMR and 13C NMR
spectroscopic analysis (vide supra): [R]20D -35.0 (c 1.03, CHCl3).
The Lewis acid-catalyzed rearrangement of R-silyloxy allylic
silanes to silyloxy vinylic silanes has been examined. The
rearrangement proceeds through an ionic mechanism involving
an ion pair. The formation of silyloxy vinylic silanes is
thermodynamically favorable due to alleviation of the steric
congestion between the two geminal silyl moieties of the
R-silyloxy allylic silane and the stabilization of the vinylic
silanes by hyperconjugation.
Experimental Section
Representative Procedure for the Isolation of R-Siloxy Allylic
Silanes. To a solution of PhMe2SiCl (1.15 equiv) in THF (0.35
M) was added finely cut lithium wire (9.2 equiv). The suspension
was stirred at 24 °C for 18 h. The resultant PhMe2SiLi solution
was transferred by cannula to a clean flask and cooled to -78 °C.
Crotonaldehyde (1.00 equiv) in THF (1.2 M) was then added
dropwise. The solution was stirred for 20 min at -78 °C, and then
the reaction mixture was added to saturated aqueous NH4Cl. The
resultant layers were separated and the aqueous layer was extracted
with Et2O. The combined organic layers were washed with saturated
aqueous NaCl. The resultant organic phase was dried over MgSO4
and filtered. The filtrate was concentrated in vacuo to afford the
R-hydroxy allylic silane as a yellow oil. The oil was then dissolved
in anhydrous DMF (2 M) followed by the addition of TBDMSCl
(1.05 equiv) and imidazole (2.05 equiv). The solution was allowed
to stir at 23 °C for a minimum of 12 h. The reaction mixture was
diluted with hexanes and the resultant solution was washed with
saturated aqueous NH4Cl. The layers were separated and the
aqueous layer was extracted with hexanes. The combined organic
layers were washed with saturated aqueous NaCl. The resultant
organic phase was dried over MgSO4 and filtered. The filtrate was
concentrated in vacuo to afford the R-silyloxy allylic silane as a
light yellow oil. Purification by flash chromatography (hexanes)
afforded the R-silyloxy allylic silanes as colorless oils.
Asymmetric Reduction with a Chiral Borane:4 R-Siloxy
Allylic Silane (-)-1. To a cooled (-78 °C) solution of (+)-DIP-
Cl (3.14 g, 9.79 mmol) in anhydrous THF (10 mL) was added a
solution of (1-oxo-2-butenyl)dimethylphenylsilane20 (0.800 g, 3.92
mmol) in anhydrous THF (10 mL). The resultant solution was then
allowed to warm to 23 °C. After the solution was stirred for 52 h
at 23 °C, diethanolamine (2.31 g, 22.0 mmol) was added. The
resultant suspension was stirred for 14 h at 23 °C. Diethyl ether
(100 mL) was added, and the solution was filtered. The filtrate
was dried over MgSO4 and filtered again. The filtrate was
concentrated in vacuo to afford the R-hydroxy allylic silane as a
light yellow oil. Purification by flash chromatography (hexanes to
90:10 hexanes:EtOAc) afforded R-hydroxy allylic silane (0.350 g,
44%) as a light yellow oil. The R-hydroxy allylic silane was isolated
in 76% ee by chiral HPLC (Chiralcel OD-H column, 99:1 hexanes/
IPA, 1 mL/min, 254 nm). The R-hydroxy allylic silane (0.165 g,
0.804 mmol) was dissolved in DMF (0.8 mL), and TBDMSCl
(0.182 g, 1.21 mmol) and imidazole (0.138 g, 2.01 mmol) were
added. The solution was stirred at 23 °C for 36 h. The reaction
mixture was diluted with hexanes (15 mL) and the resultant solution
was washed with saturated aqueous NH4Cl (15 mL). The layers
were separated and the aqueous layer was extracted with hexanes
(3 × 10 mL). The combined organic layers were washed with
saturated aqueous NaCl (15 mL). The resultant organic phase was
dried over MgSO4 and filtered. The filtrate was concentrated in
vacuo to afford the R-siloxy allylic silane as a light yellow oil.
Purification by flash chromatography (hexanes) afforded (-)-1 as
a colorless oil (0.160 g, 63%). The product was identical with (()-1
R-Silyloxy Allylic Silane 1: GC tR 12.4 min (50 °C, 10 deg/
1
min); H NMR (500 MHz, CDCl3) δ 7.55 (m, 2H), 7.35 (m, 3H),
5.42 (m, 2H), 4.04 (dt, J ) 6.1, 1.5 Hz, 1H), 1.66 (dt, J ) 6.1, 1.3
Hz, 3H), 0.88 (s, 9H), 0.30 (s, 3H), 0.28 (s, 3H), -0.04 (s, 3H),
-0.11 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 137.7, 134.6, 132.2,
129.2, 127.7, 122.2, 68.4, 26.1, 18.4, 18.0, -4.1, -5.0, -5.3, -5.6;
IR (thin film) 2957, 1472, 1252, 1084 cm-1; HRMS (ESI) m/z calcd
for C18H32OSi2Na (M + Na)+ 343.1889, found 343.1888. Anal.
Calcd for C18H32OSi2: C, 67.43; H, 10.06. Found: C, 67.23; H,
10.21.
R-Silyloxy Allylic Silane 3: GC tR 15.5 min (50 °C, 10 deg/
1
min); H NMR (500 MHz, CDCl3) δ 7.55 (m, 2H), 7.35 (m, 3H),
5.42 (m, 2H), 4.05 (m, 1H), 2.0 (m, 2H), 1.68 (septet, J ) 6.9 Hz,
1H), 0.95 (t, J ) 7.4 Hz, 3H), 0.90 (d, J ) 6.8 Hz, 3H), 0.88 (d,
J ) 6.8 Hz, 3H), 0.83 (s, 3H), 0.82 (s, 3H), 0.38 (s, 3H), 0.32 (s,
3H), 0.01 (s, 3H), -0.05 (s, 3H); 13C NMR (125 MHz, CDCl3) δ
137.7, 134.6, 130.3, 129.6, 129.2, 127.6, 68.5, 34.4, 25.7, 25.3,
20.7, 20.6, 18.9, 18.8, 14.4, -1.9, -3.1, -5.3, -5.6; IR (thin film)
2959, 1464, 1251, 1028 cm-1. Anal. Calcd for C21H38OSi2: C,
69.54; H, 10.56. Found: C, 69.65; H, 10.75.
1
by H NMR and 13C NMR spectroscopic analysis (vide supra):
[R]23 -25.1 (c 1.00, CHCl3).
D
Representative Procedure for R-Silyloxy Allylic Silane Rear-
rangement. To a cooled solution (-78 °C) of R-silyloxy allylic
silane in methylene chloride (0.1 M) was added BF3‚OEt2 (1.1
equiv). After 3 h at -78 °C an aqueous solution of NaHCO3 was
added. The resultant layers were separated and the aqueous layer
was re-extracted with EtOAc. The combined organic layers were
washed with saturated aqueous NaCl, dried over MgSO4, and
filtered. The filtrate was concentrated in vacuo to afford the silyloxy
vinylic silane as a light yellow oil. Vinylic silanes were purified
by silica gel chromatography (2% EtOAc/98% hexanes).
Asymmetric Reduction with a Chiral Lithium Amide:19
R-Silyloxy Allylic Silane (-)-1. The asymmetric reduction was
performed according to the Takeda procedure.19 To a cooled (-80
°C) solution of (1-oxo-2-butenyl)dimethylphenylsilane20 (0.600 g,
2.94 mmol) in THF (20 mL) was added chiral lithium amide
dropwise [the chiral lithium amide was freshly prepared by treating
(S)-N-(2,2-dimethylpropyl)-1-phenyl-2-(4-methylpiperazinyl)ethy-
lamine (1.02 g, 3.52 mmol) with n-butylithium (1.4 mL, 3.5 mmol,
Silyloxy vinylic silane 2: GC tR 12.8 min (50 °C, 10 deg/min);
1H NMR (500 MHz, CDCl3) δ 7.54 (m, 2H), 7.37 (m, 3H), 6.15
(dd, J ) 18.6, 4.4 Hz, 1H), 5.95 (d, J ) 18.6 Hz, 1H), 4.35 (m,
1H), 1.23 (d, J ) 6.4 Hz, 3H), 0.92 (s, 9H), 0.35 (s, 6H), 0.07 (s,
(19) Takeda, K.; Ohnishi, Y.; Koizumi, T. Org. Lett. 1999, 1, 237-
239.
(20) Danheiser, R. L.; Fink, D. M.; Okano, K.; Tsai, Y.-M.; Szczepanski,
S. W. J. Org. Chem. 1985, 50, 5393-5396.
J. Org. Chem, Vol. 72, No. 17, 2007 6597

6H); 13C NMR (125 MHz, CDCl3) δ 152.5, 139.3, 134.2, 129.3,
128.1, 124.9, 71.5, 26.3, 24.3, 18.7, -2.2, -2.3, -4.3, -4.4; IR
(thin film) 2956, 2857, 1620, 1250 cm-1; HRMS (ESI) m/z calcd
for C18H32OSi2Na (M + Na)+ 343.1889, found 343.1886.
Silyloxy vinylic silane 4: GC tR 13.7 min (50 °C, 10 deg/min);
1H NMR (500 MHz, CDCl3) δ 7.54 (m, 2H), 7.37 (m, 3H), 6.05
(dd, J ) 18.7, 5.6 Hz, 1H), 5.89 (dd, J ) 18.7, 1.0 Hz, 1H), 4.09
(m, 1H), 1.68 (septet, J ) 6.8 Hz, 1H), 1.50 (m, 2H), 0.89 (d, J )
6.8 Hz, 6H), 0.86 (m, 3H), 0.84 (s, 6H), 0.34 (s, 6H), 0.09 (s, 3H),
0.06 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 151.6, 139.2, 134.1,
129.1, 127.9, 126.3, 77.0, 34.4, 30.8, 25.3, 20.6, 18.9, 9.9, -2.0,
-2.2, -2.3, -2.5; IR (thin film) 2959, 1250, 831 cm-1; HRMS
(ESI) m/z calcd for C21H38OSi2Na (M + Na)+ 385.2359, found
385.2360. Anal. Calcd for C18H32OSi2: C, 69.54; H, 10.56.
Found: C, 69.32; H, 10.40.
separated and the aqueous layer was extracted with 3 × 5 mL of
hexanes. The combined organic layers were washed with saturated
aqueous NaCl (15 mL), dried over MgSO4, and filtered. GC analysis
revealed the formation of vinylic silanes 2 and 4 exclusively.
Acknowledgment. This research was supported by the
National Science Foundation (CHE-0315572). A.R. thanks the
National Institutes of Health for a predoctoral fellowship. A.I.K.
and K.L.K. thank the Undergraduate Research Opportunities
Program (UCI) for fellowships. K.A.W. thanks Amgen, Johnson
& Johnson, and Lilly for awards to support research. We thank
Dr. Phil Dennison for assistance with NMR spectroscopy and
Dr. John Greaves for assistance with mass spectrometry.
Supporting Information Available: Full experimental details,
spectral data, and characterization for all new compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Crossover Rearrangement. To a cooled solution (-78 °C) of
R-silyloxy allylic silane 1 (0.177 g, 0.551 mmol) and R-silyloxy
allylic silane 3 (0.200 g, 0.551 mmol) in 4 mL of solvent was added
BF3‚OEt2 (0.162 g, 1.14 mmol). After 3 h at -78 °C, an aqueous
solution of NaHCO3 (15 mL) was added. The resultant layers were
JO070634P
6598 J. Org. Chem., Vol. 72, No. 17, 2007