α-Substituted acylsilanes via a highly selective [1,4]-Wittig rearrangement of α-benzyloxyallylsilane

Article abstract of DOI:10.1039/b603228a

α-Benzyloxyallylsilane undergoes efficient [1,4]-Wittig rearrangement to generate an enolate intermediate that can be trapped with various electrophiles, thereby providing a new synthetic approach to substituted acylsilanes. The Royal Society of Chemistry

Full text of DOI:10.1039/b603228a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
a-Substituted acylsilanes via a highly selective [1,4]-Wittig
rearrangement of a-benzyloxyallylsilane{
Edith N. Onyeozili and Robert E. Maleczka Jr.*
Received (in Bloomington, IN, USA) 7th March 2006, Accepted 11th April 2006
First published as an Advance Article on the web 3rd May 2006
DOI: 10.1039/b603228a
organic synthesis, we decided to learn more about this reaction.
a-Benzyloxyallylsilane undergoes efficient [1,4]-Wittig rearran-
gement to generate an enolate intermediate that can be trapped
with various electrophiles, thereby providing a new synthetic
approach to substituted acylsilanes.
Specifically, we were interested in increasing the [1,4]/[1,2] ratio and
taking advantage of the enolate formed during the [1,4]-
sigmatropic shift. Furthermore, we envisaged that information
gathered during such a study would be helpful in future
investigations directed at mechanistic inquiries.
Wittig rearrangements of a-lithiated ethers have proven to be a
valuable tool for organic chemists.1 Among these rearrangements
the [2,3]-Wittig is certainly the most studied and synthetically
mature.1,2 Similarly, the [1,2]-Wittig rearrangement has also been
the subject of numerous mechanistic and synthetic studies,1 many
of which have come out of the labs of Nakai and Tomooka. Their
investigations,3 and those of several other groups,4 have shed
considerable light on the unique stereochemical aspects of this
radical–radical anion dissociation–recombination. Allylic ethers
are also capable of a [1,4]-Wittig rearrangement.5 Nonetheless,
relative to its [1,2]- and [2,3]-counterparts, the [1,4]-Wittig remains
a reaction with many unanswered questions. For example, whether
the [1,4]-mechanism is concerted or involves a radical–radical
anion dissociation–recombination is still debated.5c,d,f The sub-
strate scope of the [1,4]-Wittig is also not well documented and
thus its potential in synthetic organic chemistry is unclear.
Moreover, for substrates capable of both pathways, a strong
preference for [1,4] over [1,2] bond reorganization is rarely
realized,5f,g,h,i with Tomooka’s very recent report of a highly
selective [1,4]-silyl migration being a relevant exception.5j
As a general rule,1,5f,11 Wittig rearrangements are sensitive to the
base used to generate the a-lithiated ether and the temperature at
which the reaction is run. Thus these seemed reasonable variables
to examine initially during the rearrangement of 1 (Table 1).
Employing 1.5 equivalents of a 1.4 M solution of MeLi in
diethyl ether as base, compound 1 was rearranged under a variety
of temperatures. These experiments revealed that temperature
clearly affects the ratio of [1,4]- vs. [1,2]-products. Per our goal, the
[1,2]-Wittig pathway could be effectively suppressed when the
reaction temperature was kept below 260 uC. However, at this
temperature, the reaction was very slow and was incomplete after
72 h. Employing a greater excess of MeLi (3–4 equiv.) and higher
temperatures (237 uC) led to complete consumption of the starting
material; however reaction times remained long (65–72 h) and
under these conditions the [1,4]:[1,2] selectivity eroded (4:1). With
MeLi as base, the combined yield of the [1,4]- and [1,2]-products
typically averaged y68%
With these preliminary temperature studies complete, we tested
different alkylithium bases in the reaction. The results are
summarized in Table 1. n-BuLi proved to be superior to MeLi,
leading to complete conversion of the substrate (1.5 equiv. of base,
278 uC, 5 h) and affording the [1,4]-product selectively ([1,2]
During the course of an earlier study on the MeLi-promoted
Wittig rearrangements of a-alkoxysilanes,6 we found that, upon
deprotonation, a-benzyloxyallylsilane 1 rearranged to afford a
mixture of the [1,4]-Wittig product (2) and a second compound (3)
derived from the [1,2]-Wittig product,7 with acylsilane 2 favored by
a ratio of 3:1 (Scheme 1). Owing to the aforementioned questions
concerning the [1,4]-Wittig combined with recent developments by
Scheidt,8 Johnson,9 and others10 on the use of acylsilanes in
Table 1 Optimizing the [1,4]-Wittig rearrangement of 1
Base
equivalents uC
Temperature/
Yield
Time/h (%)
Entry Base
[1,4]:[1,2]
1
2
3
4
5
6
7
8
a
MeLi 1.5–2.0
18 to 20
18 to 20
280 to 237 72
280 to 237
1.0 69
1.0 68
1.4:1 to 2:1
2.45:1
4:1
n-BuLi 1.5
MeLi 3.0
n-BuLi 1.5
s-BuLi 1.5
MeLi 3.0
n-BuLi 1.5
s-BuLi 1.5
68
83
Scheme 1 Wittig rearrangement of a-alkoxysilane 1.
2
9.1:1
250 to 237 ,0.1 79–83 .20:1
.12:1
Department of Chemistry, Michigan State University, 540 Chemistry,
East Lansing, Michigan, 48824, USA.
E-mail: maleczka@chemistry.msu.edu; Fax: +1 517 353 1793;
Tel: +1 517 355 9715
{ Electronic supplementary information (ESI) available: Experimental
procedures and spectral data. See DOI: 10.1039/b603228a
280 to 250 72
280 to 275 ,5
68
79–83 .100:1a
280 to 275 ,0.5 79–83 .100:1a
[1,2]-Wittig product 3 was not detected (by TLC, 1H NMR or
GC-MS).
2466 | Chem. Commun., 2006, 2466–2468
This journal is ß The Royal Society of Chemistry 2006

View Article Online
product could not be detected by 1H NMR spectroscopy).
Allowing the reaction to warm to 237 uC afforded the [1,2] and
[1,4] products 2 and 3 in a combined 83% yield and 9:1 ratio in
favor of the [1,4] product. However, at room temperature, the
[1,4]:[1,2] selectivity was not improved over MeLi, and the yields
were comparable. s-BuLi was found to be superior to both n-BuLi
and MeLi in initiating the Wittig rearrangements of a-alkoxy-
silanes (results are shown in Table 1). Upon treatment of a cold
[1,4]-rearrangement product 2 exclusively12 and in good yield
(79–83%).13,14 To the best of our knowledge, this is the most rapid,
selective, and efficient [1,4]-Wittig rearrangement of a-alkoxysi-
lanes in particular, and allyl benzyl ethers in general, to be
reported.
We believe these data suggest different mechanisms for the [1,4]-
and [1,2]-rearrangements of 1. Previous studies on the concerted
[2,3]-Wittig determined that the stepwise [1,2]-Wittig becomes
competitive at higher temperatures.1 Thus, if concerted, a [1,4]-
reorganization should be preferred at cold temperatures, provided
the base is strong enough to deprotonate the starting material
(e.g. s-BuLi). Entries 3–6 of Table 1 are consistent with this
hypothesis. Depending on what base was employed, deprotonation
and rearrangement occurred to different extents over each
experiment’s temperature range. With weaker bases (MeLi and
n-BuLi) complete deprotonation–rearrangement only occurred
after reaction temperatures reached their upper limits and thus
more [1,2]-Wittig was seen. In contrast, s-BuLi deprotonated 1 at
the lower end of the temperature range thereby allowing the [1,4]-
Wittig to proceed nearly unopposed.
(278 uC) THF solution of our model substrate 1 with
1.5 equivalents of s-BuLi (1.3 M in cyclohexane), Wittig
rearrangement was complete in 30 min to afford the
Table 2 1,4-Wittig rearrangement–enolate trapping
Yield
(%)
Having established highly selective [1,4]-Wittig conditions, we
next sought to take advantage of the enolate generated upon
rearrangement by quenching the reaction with various electro-
philes (Table 2).15 This would establish the [1,4]-Wittig as a new
way to build a-substituted acylsilanes.
Entry Electrophile
Product(s)
1
CH2LCHCH2Br
55
As such a protocol would involve C–C bond forming reactions
at both the c- and a-carbons of the final product, the reaction
sequence would represent an alternative to the conjugate addition
of nucleophiles to 1-trimethylsilylpropenone followed by electro-
phile capture as a means of synthesizing these TMS-ketones.
Curiously enough, to the best of our knowledge, such an approach
to elaborating a,b-unsaturated acylsilanes has been used only in a
handful of specialized cases.16 As such the route described herein
appears to be unprecedented in its generality.
2
PhCH2Br
66
3
4
MeI
EtI
73
81
The results of our trapping experiments are summarized in
Table 2.17 Allylation, benzylation, and methylation afforded
only a-C-alkylated acylsilanes (5–7) in moderate to good yields
(Table 2, entries 1–3). Reaction with ethyl iodide or propyl
iodide resulted in 3:1 mixtures of the C- and O-alkylated products
(81% and 66% yields respectively) (entries 4–5). Benzaldehyde
proved a troublesome electrophile as over condensation was
difficult to control (entry 6).18 However, quenching with TMSCl
selectively gave (E)-O-silylenol ether 12 in 73% yield (entry 7).19
In light of the benzaldehyde result, the efficient generation of
the silylketene acetal is noteworthy since such compounds
react well under Mukaiyama aldol conditions to give b-alkox-
yacylsilanes.20 Similarly, enol ester 13,19 resulting from the
reaction with Ac2O, could also be obtained by this protocol
(entry 8).
5
PrI
66
6
7
PhCHO
TMSCl
Over condensation
—
73
8
9
AcCl
69
58
This process could also be used as a route to TMS-substituted
alkynes. As discovered by Fleming and Mwaniki, enol triflates of
acylsilanes are prone to rapid dehydration.21 Thus trapping with
PhNTf2 did not afford any observable amounts of the correspond-
ing vinyl triflate, but rather gave trimethyl(4-phenylbut-1-ynyl)-
silane 14 in 58% yield (entry 9). Use of the nonaflating reagent
CF3(CF2)3SO2F under similar reaction conditions resulted in the
PhNTf2
10
CF3(CF2)2SO2F
58
1
formation of the vinyl nonaflate 15 as determined from the H
NMR spectrum of the crude reaction mixture. However, even the
nonaflate proved sensitive to acidic conditions and, once subjected
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 2466–2468 | 2467

View Article Online
10 (a) C. T. Clark, B. C. Milgram and K. A. Scheidt, Org. Lett., 2004, 6,
3977–3980 and references cited therein. For reviews see: (b) P. C. B.
Page and M. J. McKenzie, in Science of Synthesis, ed. I. Fleming,
Pergamon, London, 2002, vol. 4, pp. 513–567; (c) B. F. Bonini,
M. C. Franchini and M. Fochi, Gazz. Chim. Ital., 1997, 127, 619–628;
(d) P. F. Cirillo and J. S. Panek, Org. Prep. Proced. Int., 1992, 24,
553–582.
11 H. Felkin and A. Tambute, Tetrahedron Lett., 1969, 10, 821–822.
12 [1,2]-Wittig product 3 was not detected (by TLC, 1H NMR or
GC-MS).
13 Wittig rearrangement of compound 1: a solution of 76 mg (0.34 mmol)
of a-alkoxysilane 1 in 4.5 mL of freshly distilled dry THF, was cooled to
278 uC under N2. s-BuLi (1.3 M in cyclohexane, 0.4 mL, 0.52 mmol,
1.5 equiv.) was added dropwise via syringe. The reaction mixture was
stirred for 30 min at 278 uC and then quenched with saturated aqueous
NH4Cl, diluted with diethyl ether, and subsequently washed with H2O
and brine. The organic phase was dried over MgSO4 and concentrated.
Silica gel chromatography (0–2% EtOAc–hexane gradient) afforded
63 mg (82%) of acylsilane 2 as a light yellow oil. IR (neat) 2955, 1717,
1643, 1497, 1454, 1250 cm21; 1H NMR (500 MHz, CDCl3) d 7.27–7.17
(m, 5H), 2.62–2.59 (overlapping dd, J = 7.3, 6.8 Hz, 2H), 2.58–2.55
(overlapping dd, J = 7.8, 7.3 Hz, 2H), 1.87–1.81 (m, 2H), 0.16 (s, 9H);
13C NMR (125 MHz, CDCl3) d 248.1, 141.8, 128.4, 128.3, 125.8, 47.5,
35.2, 23.6, 23.2. HRMS (EI) m/z 219.1210 [(M 2 H)+; calcd for
C13H19OSi 219.1210].
to silica gel column chromatography, it too underwent elimination
to give 14 in the same 58% isolated yield (entry 10).
In summary, we have established that, upon deprotonation with
s-BuLi, a-benzyloxyallylsilane (1) undergoes [1,4]-Wittig rearran-
gement with unprecedented selectivity. By concluding the reaction
with the addition of an electrophile, a-benzyloxyallylsilane serves
as a unique source of a variety of a-substituted acylsilanes.
We thank Feng Geng for preliminary studies as well as the NIH
(HL-58114), NSF (CHE-9984644), and the Astellas USA
Foundation for their generous support.
Notes and references
1 (a) T. Nakai and K. Tomooka, Pure Appl. Chem., 1997, 69, 595–600; (b)
T. Nakai and K. Mikami, Org. React., 1994, 46, 105–209; (c)
K. Tomooka, in Chemistry of Organolithium Compounds, ed.
Z. Rappoport and M. Ilan, Wiley, London, 2004, vol. 2, pp. 749–828.
2 For recent examples see: (a) M. Sasaki, M. Higashi, H. Masu,
K. Yamaguchi and K. Takeda, Org. Lett., 2005, 7, 5913–5915; (b)
G. McGowan, Aust. J. Chem., 2002, 55, 799; (c) J. C. Anderson,
A. Flaherty and M. E. Swarbrick, J. Org. Chem., 2000, 65, 9152–9156;
(d) S. A. Hart, C. O. Trindle and F. A. Etzkorn, Org. Lett., 2001, 3,
1789–1791; (e) R. E. Maleczka, Jr. and F. Geng, Org. Lett., 1999, 1,
1111–1113; (f) P. Antoniotti and G. Tonachini, J. Org. Chem., 1998, 63,
9756–9762; (g) S. E. Gibson, P. Ham and G. R. Jefferson, Chem.
Commun., 1998, 123–124.
14 For alternative preparations of 2 see: (a) N. G. Bhat, A. Tamm and
A. Gorena, Synlett, 2004, 297–298; (b) C. Hammaecher,
I. Ouzzane, C. Portella and J.-P. Bouillon, Tetrahedron, 2005, 61,
657–663.
15 Enolate 4 had been trapped previously with allyl bromide to afford
a-allyl acylsilane 5 in 15% yield. F. Geng, PhD thesis, Michigan State
University, 2001.
3 (a) K. Tomooka, T. Igarashi and T. Nakai, Tetrahedron Lett., 1993, 34,
8139–8142; (b) K. Tomooka, T. Igarashi and T. Nakai, Tetrahedron,
1994, 50, 5927–5932; (c) K. Tomooka, H. Yamamoto and T. Nakai,
J. Am. Chem. Soc., 1996, 118, 3317–3318; (d) K. Tomooka,
H. Yamamoto and T. Nakai, Liebigs Ann., 1997, 1275–1281; (e)
K. Tomooka, H. Yamamoto and T. Nakai, Angew. Chem., Int. Ed.,
1999, 38, 3741–3743.
4 (a) S. L. Schreiber and M. T. Goulet, Tetrahedron Lett., 1987, 28,
1043–1046; (b) E. J. Verner and T. Cohen, J. Am. Chem. Soc., 1992, 114,
375–377; (c) R. Hoffmann and R. Bru¨ckner, Chem. Ber., 1992, 125,
1957–1963; (d) R. E. Maleczka, Jr. and F. Geng, J. Am. Chem. Soc.,
1998, 120, 8551–8552 and references cited therein.
5 (a) U. Scho¨llkopf, Angew. Chem., Int. Ed. Engl., 1970, 9, 763–773; (b)
U. Scho¨llkopf, K. Fellenberger and M. Rizk, Justus Liebigs Ann. Chem.,
1970, 734, 106–115; (c) H. Felkin and C. Frajeermann, Tetrahedron
Lett., 1977, 18, 3485–3488 and references cited therein; (d) K. Sayo,
Y. Kumara and T. Nakai, Tetrahedron Lett., 1982, 23, 3931–3934; (e)
K. Hayakawa, A. Hayashida and K. Kanematsu, J. Chem. Soc., Chem.
Commun., 1988, 1108–1110; (f) M. Schlo¨sser and S. Strunk,
Tetrahedron, 1989, 45, 2649–2664; (g) W. F. Bailey and L. M. Zarcone,
Tetrahedron Lett., 1991, 32, 4425–4426; (h) W. F. Bailey, E. R. Punzalan
and L. M. Zarcone, Heteroat. Chem., 1992, 3, 55–61; (i) K. Tomooka,
H. Yamamoto and T. Nakai, Angew. Chem., Int. Ed., 2000, 39,
4500–4502; (j) A. Nakazaki, T. Nakai and K. Tomooka, Angew. Chem.,
Int. Ed., 2006, 45, 2235–2238.
16 (a) a,b-Unsaturated silylketones have been reported to undergo
conjugate additions with cuprates, however no details of these reactions
were described in: H. J. Reich, M. J. Kelley, R. E. Olson and
R. C. Holtan, Tetrahedron, 1983, 39, 949–960. For annulations with
allenylsilanes see: (b) R. L. Danheiser and D. M. Fink, Tetrahedron
Lett., 1985, 26, 2513–2516. For cyclopropanations see: (c) J. S. Nowick
and R. L. Danheiser, Tetrahedron, 1988, 44, 4113–4134. For examples
of 1,4-additions with TMS-substituted nucleophiles see: (d) A. Ricci,
A. Degl’Innocenti, G. Borselli and G. Reginato, Tetrahedron Lett., 1987,
28, 4093–4096; (e) A. Degl’Innocenti, P. Ulivi, A. Capperucci,
G. Reginato, A. Mordini and A. Ricci, Synlett, 1992, 883–886. Also
see: (f ) I. A. Stergiades and M. A. Tius, J. Org. Chem., 1999, 64,
7547–7551.
17 Electrophilic trapping of enolate 5: the Wittig rearrangement was carried
out as described above (ref. 13). Upon complete rearrangement (ca.
30 min. at 278 uC), the resultant enolate solution was transferred via
cannula to a THF solution of the electrophile at the indicated
temperature (Table 2). The reaction was then stirred for the indicated
time with monitoring by TLC. The mixture was then quenched with
saturated aqueous NH4Cl and diluted with diethyl ether. Phases were
separated and the organic phase washed with H2O and brine. Workup
was then carried out as described above to afford the a-substituted
acylsilane.
18 NMR and HRMS analysis suggest two molecules of benzaldehyde are
incorporated in the product, but the exact structure of this compound is
unclear.
19 The E geometries of the both 12 and 13 were based on the spectral data
for 13 being identical to those previously reported for that geometric
isomer. See: M. Yamane, K. Uera and K. Narasaka, Bull. Chem. Soc.
Jpn., 2005, 78, 477–486.
20 (a) M. Honda, W. Oguchi, M. Segi and T. Nakajima, Tetrahedron,
2002, 58, 6815–6823; (b) S. Itsuno, S. Arima and N. Haraguchi,
Tetrahedron, 2005, 61, 12074–12080.
21 I. Fleming and J. M. Mwaniki, J. Chem. Soc., Perkin Trans. 1, 1998,
1237–1247.
6 (a) R. E. Maleczka, Jr. and F. Geng, Org. Lett., 1999, 1, 1115–1118; (b)
for the preparation of 1 see: ref. 2d.
7 In ref. 6a, we mistakenly assigned 3 as a [2,3]-Wittig derived product. In
actuality, this product is the result of a [1,2]-Wittig followed by
migration of the TMS group to the b-carbon. If this migration is the
result of a direct [1,3]-shift or involves a Brook/vinylogous retro-Brook
rearrangement is unclear at this time (also see ref. 5j).
8 (a) A. E. Mattson and K. A. Scheidt, Org. Lett., 2004, 6, 4363–4366; (b)
A. E. Mattson, A. R. Bharadwaj and K. A. Scheidt, J. Am. Chem. Soc.,
2004, 126, 2314–2315.
9 (a) D. A. Nicewicz, C. M. Yates and J. S. Johnson, J. Org. Chem., 2004,
69, 6548–6555; (b) X. Linghu, D. A. Nicewicz and J. S. Johnson, Org.
Lett., 2002, 4, 2957–2960.
2468 | Chem. Commun., 2006, 2466–2468
This journal is ß The Royal Society of Chemistry 2006