A one-pot silyl-Reformatsky olefination

Article abstract of DOI:10.1016/j.tetlet.2007.10.029

A novel one-pot olefination reaction has been developed, involving the stereoselective formation of (E)-α,β-unsaturated esters/ketones from the reaction of α-bromocarbonyl compounds with aromatic aldehydes. The reactions use a reagent combination of trichlorosilane, and triethylamine and may proceed via the in situ formation of a trichlorosilyl ketene acetal. The general procedure offers key advantages (high conversions, low quantities of organic soluble by-products) over the conventional Wittig reaction.

Full text of DOI:10.1016/j.tetlet.2007.10.029

Available online at www.sciencedirect.com
Tetrahedron Letters 48 (2007) 8687–8690
A one-pot silyl-Reformatsky olefination
James M. Smith and Michael F. Greaney*
University of Edinburgh, School of Chemistry, Joseph Black Building, King’s Buildings, West Mains Road, Edinburgh EH9 3JJ, UK
Received 3 September 2007; revised 25 September 2007; accepted 4 October 2007
Available online 9 October 2007
Abstract—A novel one-pot olefination reaction has been developed, involving the stereoselective formation of (E )-a,b-unsaturated
esters/ketones from the reaction of a-bromocarbonyl compounds with aromatic aldehydes. The reactions use a reagent combination
of trichlorosilane, and triethylamine and may proceed via the in situ formation of a trichlorosilyl ketene acetal. The general proce-
dure offers key advantages (high conversions, low quantities of organic soluble by-products) over the conventional Wittig reaction.
ꢀ
2007 Elsevier Ltd. All rights reserved.
1
Regioselective olefination methods such as the Wittig,
olefination reactions, which benefit from easier purifica-
tion methods remain attractive research targets.
2
3
Julia–Lythgoe and Peterson reactions lie at the heart
of organic chemistry and are indispensable for the con-
struction of the carbon–carbon double bond. Since its
discovery in 1953, the Wittig olefination in particular
remains extremely popular in synthetic organic chemis-
try. The reaction features broad substrate scope, excel-
lent functional group tolerance and high yields—
frequently accompanied by good levels of discrimination
between E and Z double bond geometric isomers in the
olefin products. However, the drawbacks of the reaction
We now report a novel olefination reaction, using the
common Wittig substrates of a-bromocarbonyl com-
pounds and aromatic aldehydes, to give the correspond-
ing (E )-a,b-unsaturated carbonyl products in moderate
to good yields in a one-pot transformation. Our work
began by studying the reaction of methyl a-bromoace-
tate 1 with anisaldehyde 2, using a reagent combination
of trichlorosilane and triethylamine. The corresponding
olefin, 3, was produced in a 39% yield (Table 1, entry 1).
A screen of solvents and other variables revealed that
the reaction performed better at lower temperatures
(À10 ꢁC) and using CH Cl as a solvent, raising the yield
4
in terms of atom economy are well-documented—the
generation of 1 equiv of triphenylphosphine oxide during
the reaction, an organic soluble by-product, which usu-
ally requires separation from the olefin product, presents
a major purification issue, especially when the reaction is
carried out on an industrial scale. Many creative solu-
tions have been developed to overcome this problem
2
2
to 59% (Table 1, entry 2).
Further optimisation centred on the use of Lewis basic
additives in the reaction, several of which increased
5
within the context of the Wittig reaction, but alternative
Table 1. Olefination of anisaldehyde with methyl bromoacetate
O
O
O
Br
+
H
OMe
OMe
MeO
MeO
1
2
3
Entry
Reagents
Solvent/temperature
MeCN/rt
Additive
Yield (%)
1
2
3
Cl
3
Cl
3
Cl
3
SiH (1.5 equiv) NEt
SiH (2 equiv) NEt
SiH (2 equiv) NEt
3
(5 equiv)
(5 equiv)
(5 equiv)
—
—
39
59
69
3
CH
CH
2
Cl
Cl
2
/À10 ꢁC to rt
3
2
2
/À10 ꢁC to rt
3
Ph PO (0.1 equiv)
*
0
Corresponding author. Tel.: +44 131 6504726; fax: +44 131 6504743; e-mail: Michael.Greaney@ed.ac.uk
040-4039/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.10.029

8688
J. M. Smith, M. F. Greaney / Tetrahedron Letters 48 (2007) 8687–8690
the yield, the best being a catalytic quantity of triphen-
ylphosphine oxide. Under these conditions, product 3
could be isolated in a 69% yield (Table 1, entry 3).
was also a suitable substrate, allowing access to the
a,b,c,d-unsaturated ester 3i, although in this case the
yield was low (Table 2, entry 9).
The scope of the reaction was explored using a variety of
aromatic aldehydes (Table 2). Yields were found to be
higher in general for benzaldehydes bearing electron-
releasing substituents on the aromatic ring (Table 2,
entries 1, 3 and 6). An ortho-methyl group was well
tolerated on the aromatic ring of the aldehyde (Table
In all cases, the olefin was produced exclusively as the E
isomer, with none of the Z isomer detected in any case,
as established by H NMR analysis of the crude reaction
1
mixtures. The reaction could also be demonstrated effi-
ciently in the case of a-bromoketones 4—yields in these
cases based on a short series were consistently high
(Table 3). Attempts to extend this methodology further
to aliphatic aldehydes and to ketones have been largely
unsuccessful at this point in time, with only very small
quantities of the desired olefin products detected in
representative cases.
2
, entry 6), in contrast to the aromatic nitro functional-
ity, which showed incompatibility towards the reaction
conditions (Table 2, entry 7)—a very low yield of the
desired product was observed. 2-Naphthaldehyde
reacted as desired to give the corresponding olefin in a
reasonable yield (Table 2, entry 8), and cinnamaldehyde
A reasonable mechanistic pathway for the general reac-
tion is presented in Scheme 1. The a-bromocarbonyl
compound 6 can react with the trichlorosilane in
the presence of triethylamine, leading to the in situ
formation of a silyl ketene acetal/silyl enol ether 7 in a
Table 2. Olefination of substituted benzaldehydes
Cl SiH (2 equiv),
3
NEt3 (5 equiv)
O
6
PPh O (0.1 equiv)
3
‘silyl-Reformatsky’-type process. The synthesis of tri-
chlorosilylenol species through the direct reaction of
1
+
RCHO
a-i
R
OMe
CH Cl2, -10 °C to r.t.
2
2
3a-i
trichlorosilane with a-halo-carbonyls is known for both
a
7
8,9
Entry
Aldehyde
Product yield (%)
a-bromoesters and a-bromoketones.
Compound 7
may then react with the aldehyde to give (either) b-(tri-
chlorosilyloxy)-carbonyl 8 or b-hydroxy-carbonyl 9 (the
O
H
1
2
3
4
5
6
7
8
9
69
MeO
Table 3. Olefination of bromoketones
O
Cl SiH (2 equiv),
3
O
3
NEt (5 equiv)
54
59
52
50
55
12
58
36
H
O
PPh O (0.1 equiv)
R
3
+
2
Br
R
CH Cl , -10 °C to r.t.
MeO
2
2
4
5a-c
O
MeO
MeO
Entry
1
Bromoketone
Product yield (%)
73
H
O
Br
O
H
O
Cl
F
Br
2
3
71
70
O
OMe
Cl
H
O
Br
O
H
Me
O
O
O
R
Br
H
R
Ar
6
10
O N
2
O
SiCl H,
Et N
3
3
-
HOSiCl3
- H O
2
H
O
H
+
H O or
2
HOSiCl
3
OH
O
O
OSiCl3
Ar
SiCl O
O
3
Ar
R
H
R
Ar
R
7
8
or
9
a
Isolated yields after column chromatography.
Scheme 1. Proposed reaction pathway.

J. M. Smith, M. F. Greaney / Tetrahedron Letters 48 (2007) 8687–8690
8689
former may be converted to the latter under the reaction
further reactions with the silane to become integrated
into the polymeric material).
1
0
conditions). Elimination of either Cl SiOH (in the case
3
of 8) or H O (in the case of 9) leads to the observed
2
olefin product 10.
In summary, a novel one-pot olefination reaction has
been developed, producing E-olefins from a-bromo-
carbonyl compounds and aldehydes with high stereo-
In certain reactions, the presence of the b-hydroxy-
carbonyl intermediates 9 was detected in the crude H
1
selectivity, using the reagent combination of Cl SiH/
3
NMR spectrum following work-up (the intermediate
made up less than 10% of the yield in such reactions,
as calculated by H NMR integrals). In order to further
confirm the propensity for this latter dehydration step,
b-hydroxy-ester 9 was treated under the reaction condi-
tions, and indeed the olefin product was detected, albeit
in low conversion (Scheme 2). It is possible that the b-
hydroxy-ester is formed in the olefination reaction itself
mainly as the trichlorosilyl ether 8, and that elimination
for this compound is faster than for 9.
NEt . Many aromatic aldehydes were found to be com-
3
patible with the reaction, which may proceed via the
in situ generation of a trichlorosilyl enol ether or ketene
acetal. Product yields were found to increase by the use
1
of sub-stoichiometric quantities of Ph PO. Practically,
3
this reaction represents a useful alternative to the stan-
dard Wittig procedure. Further exploration of the scope
of the reaction with a wider variety of substrate classes
and additives will be the focus of future work in this
area.
When using an alternative base system in the reaction
i
(
1.8 equiv of Pr NEt), a mixture of b-hydroxy-ester 9
Acknowledgement
2
and olefin 3a was recovered, lending further credence
to this mechanism (Scheme 3). The effect of triphenyl-
phosphine oxide additive remains unclear, though it is
possible that it may be acting as a Lewis base to activate
the silyl ketene acetal towards electrophilic attack by the
aldehyde, in analogy to Denmark’s aldol chemistry of
We acknowledge the EPSRC for financial support.
References and notes
1
0
1. (a) Wittig, G.; Geissler, G. Liebigs Ann. Chem. 1953, 580,
chlorosilyl enolates. Studies showed that there was
little effect on product yields upon raising the loading
4
8
4–57; (b) Maryanoff, B. E.; Reitz, A. B. Chem. Rev. 1989,
9, 863–927; (c) Murphy, P. J.; Lee, S. E. J. Chem. Soc.,
1
1
of the phosphine oxide beyond the 10 mol % level.
Perkin Trans. 1 1999, 3049–3066.
2
3
. (a) Julia, M.; Paris, J. M. Tetrahedron Lett. 1973, 14,
In practical terms, the reaction is very easy to carry out,
and the work-up is straightforward (the silane by-prod-
ucts are polymeric and can be removed easily by simple
4
833–4836; (b) Julia, M. Pure Appl. Chem. 1985, 57, 763–
768; (c) Kocienski, P. J.; Lythgoe, B.; Ruston, S. J. Chem.
Soc., Perkin Trans. 1 1978, 829; (d) Blakemore, P. J.
Chem. Soc., Perkin Trans. 1 2002, 2563–2585.
. (a) Peterson, D. J. J. Org. Chem. 1968, 33, 780–784; (b)
van Staden, L. F.; Gravestock, D.; Ager, D. J. Chem. Soc.
Rev. 2002, 31, 195–200.
1
2
filtration following aqueous quenching). In the case
of the bromoester reaction, the large majority of the
organic-soluble residue was the desired product (as
1
observed by H NMR of the crude material). In the case
4
5
. Trost, B. M. Science 1991, 254, 1471–1477.
of the bromoketone reaction, in addition to the desired
product, some debrominated ketone was also detected—
this material is presumably formed by protonation of
the trichlorosilyl enol ether intermediate. The reaction
goes with full conversion with respect to the aromatic
aldehydes (given the yields, a minor percentage of the
aldehyde material may be lost through reduction by
the silane—the resultant alcohol would then undergo
. For examples of the use of solid-supported phosphines,
perfluorinated phosphines and water-soluble phosphines
in the Wittig reaction, see: (a) Westman, J. Org. Lett.
2
001, 3, 3745–3747; (b) Galante, A.; Lhoste, P.; Sinou, D.
Tetrahedron Lett. 2001, 42, 5425–5427; (c) Russell, M. G.;
Warren, S. Tetrahedron Lett. 1998, 39, 7995–7998.
6. (a) Latouche, R.; Texier-Boullet, F.; Hamelin, J. Tetra-
hedron Lett. 1991, 32, 1179–1182; (b) Latouche, R.;
Texier-Boullet, F.; Hamelin, J. Bull. Soc. Chim. Fr. 1993,
1
30, 535–546.
7
. Inoue, Y.; Arai, M.; Miyoshi, T. Japanese Patent JP
03209387 A 19910912 1991; Chem. Abstr. 1991, 116,
83923.
Cl SiH (2 equiv),
3
NEt3 (5 equiv)
OH
O
PPh O (0.1 equiv)
3a (15%)
3
8. Benkeser, R. A.; Smith, W. E. J. Am. Chem. Soc. 1968, 90,
307–5309.
9. (a) Denmark, S. E.; Stavenger, R. A.; Wong, K.-T.
Tetrahedron 1998, 54, 10389–10402; (b) Denmark, S. E.;
Stavenger, R. A.; Winter, S. B. D.; Wong, K.-T.; Barsanti,
P. A. J. Org. Chem. 1998, 63, 9517–9523.
OMe
+
5
9
(75%)
CH Cl , -10 °C to r.t.
2
2
MeO
9
Scheme 2. Dehydration of aldol product.
1
0. (a) Denmark, S. E.; Winter, S. B. D. Synlett 1997, 9, 1087–
1
089; (b) Denmark, S. E.; Wong, K. T.; Stavenger, R. A.
Cl SiH (1.8 equiv)
Pr EtN (6 equiv)
2
3
J. Am. Chem. Soc. 1997, 119, 2333–2334; (c) Denmark, S.
E.; Stavenger, R. A. Acc. Chem. Res. 2000, 33, 432–440;
i
1
+
2
9 (45%)
+
3a (18%)
(
d) Denmark, S. E.; Fan, Y. J. Am. Chem. Soc. 2002, 124,
Ph PO (0.1 equiv)
4233–4235; (e) Denmark, S. E.; Fan, Y.; Eastgate, M. D.
J. Org. Chem. 2005, 70, 5235–5248.
3
o
CH Cl , -10 C
2
2
1
3
1. The possibility of the Ph PO additive ameliorating the
i
2
Scheme 3. Formation of b-hydroxy-ester 9 using Pr NEt.
yield via a minor Wittig reaction pathway was considered,

8690
J. M. Smith, M. F. Greaney / Tetrahedron Letters 48 (2007) 8687–8690
as Cl
3
SiH/NEt
3
might be expected to reduce Ph
3
PO to
4 mmol), which was added in one portion. The resultant
mixture was stirred at À10 ꢁC for 8 h, then gradually
warmed to rt overnight. The reaction was quenched by
Ph P under the reaction conditions. Control experiments
3
indicate that Ph PO does indeed undergo slow reduction
3
to the phosphine under the reaction conditions. However,
when the olefination of anisaldehyde was carried out using
careful addition of satd aq NaHCO
3
(10 mL) followed by
H
2
O (20 mL). Et O (40 mL) was added, and the precip-
2
10 mol % of Ph
3
P, a lower isolated yield of 48% of 3a was
itate removed by filtration. The aqueous phase was further
obtained (compared to 59% obtained in the absence of
additives, Table 1, entry 2). This result suggests that either
extracted with Et O (2 · 30 mL), the combined organic
2
4
extracts were washed with brine, dried (MgSO ), and
the Wittig pathway is not contributing at all, or that Ph
3
P
concentrated in vacuo. The crude product was purified by
inhibits the silyl-Reformatsky pathway to a greater extent
than the promotion of the Wittig pathway, to give an
overall lower yield. On the basis of these and other
experimental observations, it appears unlikely the Wittig
pathway is operational.
flash chromatography on silica gel (hexane/Et O, 3:1)
2
1
3
providing methyl (E)-3-(4-methoxyphenyl)propenoate
1
3a (265 mg, 69%) as a white crystalline solid. H NMR
(250 MHz, CDCl ): d 3.80 (3H, s, OCH ); 3.84 (3H, s, Ar-
OCH ); 6.31 (1H, d, J = 16.0, CHCO Me); 6.91 (2H, d,
3
3
3
2
1
2. A typical reaction is carried out as follows: To a stirred
solution of triphenylphosphine oxide (56 mg, 0.2 mmol)
and anisaldehyde (272 mg, 2 mmol) in anhydrous CH Cl
6 mL) under N at À10 ꢁC was added methyl a-bromo-
J = 8.9, Ar-H); 7.48 (2H, d, J = 8.9, Ar-H); 7.66 (1H, d,
13
2 3
J = 16.0, CHCHCO Me); C NMR (62.5 MHz, CDCl ):
2
2
d 51.53, 55.34, 114.34, 115.28, 127.12, 129.74, 144.52,
161.43, 167.72.
13. Zhang, Z.; Wang, Z. J. Org. Chem. 2006, 71, 7485–
(
2
acetate (704 mg, 4.6 mmol), followed by freshly distilled
NEt (1.01 g, 10 mmol), and finally Cl SiH (541 mg,
3
3
7487.