One-pot sequential Baylis-Hillman and Michael reactions

Article abstract of DOI:10.1016/j.tetlet.2004.07.122

A new one-pot procedure for the sequential Baylis-Hillman and Michael reactions has been developed to construct two carbon-carbon bonds with three components. This procedure has been applied to combine a variety of aromatic aldehydes, β-unsubstituted acry

Full text of DOI:10.1016/j.tetlet.2004.07.122

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 7141–7143
One-pot sequential Baylis–Hillman and Michael reactions
Wengui Wang* and Marvin Yu
Department of Synthetic Chemistry, GlaxoSmithKline Pharmaceuticals, 709 Swedeland, PO Box 1539, King of Prussia,
PA 19406, USA
Received 26 January 2004; revised 21 July 2004; accepted 21 July 2004
Available online 14 August 2004
Abstract—A new one-pot procedure for the sequential Baylis–Hillman and Michael reactions has been developed to construct two
carbon–carbon bonds with three components. This procedure has been applied to combine a variety of aromatic aldehydes,
b-unsubstituted acrylates and activated methide nucleophiles to generate highly diversified and functionalized organic compounds.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Practical and efficient construction of highly functional-
ized and diversified molecules from readily available raw
materials is a great challenge and highly desirable.
Assembly of multiple components in a one-pot opera-
tion represents one of the possible avenues.1 During
our investigation for a practical synthesis of a develop-
mental calcium receptor antagonist, we sought to trans-
form the three readily available and inexpensive
components of benzaldehyde, methyl acrylate and 2-
nitropropane into molecule 1 (Scheme 1). We envisioned
that the sequential Baylis–Hillman2 and Michael reac-
tions would represent a rapid and economical way to
assemble these three components. To the best of our
knowledge, there are no documented methods in the lit-
erature for one-pot, sequential Baylis–Hillman and
Michael reactions to form two carbon–carbon bonds
from three separate components. We report here our
investigations on the one-pot, sequential Baylis–Hillman
and Michael reactions and its application for the rapid
construction of highly functionalized and diversified
organic molecules from simple raw materials.
The Baylis–Hillman reaction is a superb example of an
atom economic3 and green4 reaction. It has drawn con-
siderable attention in the last few years due to its atom
economy, potential catalytic nature, mild reaction con-
ditions, toleration of oxygen and water as well as com-
patibility of diverse functionalities. In practice, the
Baylis–Hillman reaction has several issues that have
needed to be addressed, such as the use of stoichiometric
instead of catalytic amounts of base, low yield and very
long reaction times that range from days to weeks. More
recent research on catalysts (Lewis base or Lewis acid),
solvent systems and asymmetric catalysis has shined the
light on the future development of this useful reaction.5
Aggarwal and Mereu6 has demonstrated that 1,8-diaza-
bicyclo[5.4.0]undec-7ene (DBU) is a far superior catalyst
to the more commonly used base 1,4-diazabicyclo[2,2,2]-
octane (DABCO). Fiftyfold rate accelerations were
achieved when employing DBU instead of DABCO.
Based on these observations, DBU was chosen as base
catalyst for our studies.
OH
OH
O
NO2
+
NO2
CO2Me
CO2Me
CO2Me
1
Scheme 1.
Keywords: Baylis–Hillman; Michael; Sequential; One-pot; DBU; Multi-component.
*
Corresponding author. E-mail: wengui.2.wang@gsk.comc
Present address: Fluorous Technologies, Inc., UPARC 970 William Pitt Way, Pittsburgh, PA 15238, USA.
0040-4039/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.07.122

7142
W. Wang, M. Yu / Tetrahedron Letters 45 (2004) 7141–7143
Considering the efficiencies of the Baylis–Hillman reac-
tion, we reasoned that a catalytic, sequential Baylis–
Hillman and Michael reaction without solvent would
be an ideal atom economical and green process. To this
end, we initially conducted our experiments at room
temperature under solvent-free condition with DBU as
the base catalyst. Thus, methyl acrylate was added to
a mixture of aldehyde and DBU with evidence of a small
exotherm and a formation of a viscous material. The
Baylis–Hillman step proceeded smoothly and was com-
plete in less than 5h, as monitored by TLC and HPLC.
Once complete, 2-nitropropane was added slowly to
accomplish the subsequent Michael addition. Unfortu-
nately, this step was highly exothermic and resulted in
an even more viscous material which could not be prop-
erly stirred, thus causing local overheating. Although
solvent-free conditions would be ideal, it was demon-
strated to be unsafe and impractical. Simply adding a
small amount of THF made the process more practical
and eliminated the safety issues originating from the
highly exothermic nature of the Michael reaction. How-
ever, addition of THF as solvent slowed the rate of the
Baylis–Hillman reaction such that it now typically
needed 2–24h to achieve >95% conversion.7 Even still,
under our typical reaction conditions, the one-pot
sequential Baylis–Hillman/Michael sequence afforded
the desired product 1 in 60% isolated yield (Scheme 2).
This multi-component procedure was successfully ex-
tended to a variety of functionalized aromatic aldehydes
and nucleophiles. All reactions gave the desired prod-
ucts in good yields except for 2-bromo-benzaldehyde
(26%), as shown in Scheme 3.8 Roughly 50:50 mixtures
of two diastereoisomers (syn/anti) were obtained in all
reactions except for 9, where an even mixture of all four
diastereoisomers was produced, as determined by HPLC
and LC–MS of the reaction mixture and 1H NMR of the
crude product. In addition, it was intriguing that one of
the two diastereoisomers of molecules 7 and 10 was
apparently unstable and disappeared during workup
and silica chromatography. Only one pure diastereo-
isomer was obtained in 26% and 34% yield with 7 and
10, respectively, although equal amounts of both
diastereoisomers were detected while monitoring the
reaction by HPLC, LC–MS and 1H NMR. When 1-
nitrocyclohexene was employed as the nucleophile, the
Michael reaction proceeded by allylic deprotonation,
followed by selective reaction at the a-carbon, to afford
compound 9 as an even mixture of four diastereoisomers
in 68% yield after silica chromatography.
It is worth pointing out that nucleophiles with more
than one acidic proton, such as nitromethane and Mel-
drumÕs acid, gave complex reaction mixtures. With b-
substituted acrylates, such as methyl crotonate, methyl
OH
OH
O
NO2
DBU
+
NO2
CO2Me
CO2Me
CO2Me
THF
1 (60% isolated yield)
Scheme 2.
OH
1. DBU/THF
2. H-Nu
Ar CHO
Nu
Ar
+
CO2R
R = Me, t-Bu
CO2R
2-10
OH
OH
OH
NO2
NO2
NO2
CO2t-Bu
CO2Me
CO2Me
N
Br
4 (26%)
NO2 OH
3 (51%)
2 (60%)
NO2 OH
NO2
OH
Me
NO2
CO2Me
CO2Me
NO2
CO2Me
CO2Me
CO2Me
5 (62%)
6 (60%)
7 (26%)
Single diastereomer
OH
OH
Me
OH
NO2
CO2Me
CO2Me
NO2
CO2Me
CO2Me
8 (55%)
CO2Me
Br
Br
Br
9 (68%)
10 (34%)
Single diastereomer
Scheme 3.

W. Wang, M. Yu / Tetrahedron Letters 45 (2004) 7141–7143
7143
Chem. Rev. 1996, 96, 1–600; (c) Bienayme, H.; Hulme, C.;
Oddon, G.; Schmitt, P. Chem. Eur. J. 2000, 6, 3321; (d)
Trost, B. M. Acc. Chem. Res. 2002, 35, 695–705; (e) Ho,
T.-L. Tandem Organic Reactions; John Wiley & Sons: New
York, 1992.
NO2
O
DBU
THF
1
NO2
+
+
+
MeO2C
90% yield
MeO2C
<5%
2. For historical development of Baylis–Hillman reaction:
Drewes, S. E.; Roos, G. H. P. Tetrahedron 1988, 44,
4653–4670.
Scheme 4.
3. (a) Trost, B. M. Science 1991, 254, 1471–1477; (b) Trost
B. M. Angew. Chem., Int. Ed. 1995, 34, 259–281.
4. For topics on the green chemistry, see two special issues: (a)
Pure Appl. Chem. 2000, 72(7); (b) Pure Appl. Chem. 2001,
73(8).
trans-cinnamate and ethyl 3,3-dimethyl acrylate, unfor-
tunately, no Baylis–Hillman reaction occurred. Using
the less reactive aldehyde 4-methoxybenzaldehyde gave,
not surprisingly, no Baylis–Hillman product under these
conditions. With DABCO as catalyst, only 2-nitro-benz-
aldehyde gave 60% yield of the desired Baylis–Hillman
and Michael adduct; the other aromatic aldehydes
stopped after the Baylis–Hillman reaction and afforded
no Michael adducts. The reverse sequential sequence,
Michael addition then Aldol condensation (i.e., DBU,
methyl acrylate and 2-nitropropane were mixed first,
and the aldehyde introduced later), gave no substantial
desired product 1 (<5%). Instead, the Michael addition
product was isolated in 90% yield (Scheme 4).
5. For reviews on Baylis–Hillman reaction: (a) Basavaiah, D.;
Rao, P. D.; Hyma, R. S. Tetrahedron 1996, 52, 8001–8062;
(b) Ciganek, E. Org. React. 1997, 51, 201–350; (c) Langer,
P. Angew. Chem., Int. Ed. 2000, 39, 3049–3052; (d)
Basavaiah, D.; Rao, A. J.; Satyanarayana, T. Chem. Rev.
2003, 103, 811–892; For recent studies on Baylis–Hillman:
(e) Aggarwal, V. K.; Emme, I.; Fulford, S. Y. J. Org. Chem.
2003, 68, 692–700; (f) Dunn, P.; Hughes, M. L.; Searle, P.
M.; Wood, A. S. Org. Process Res. Dev. 2003, 7, 244–253;
(g) Shi, M.; Xu, Y.-M. J. Org. Chem. 2003, 68, 4784–4790;
(h) Karur, S.; Hardin, J.; Headley, A.; Li, G. Tetrahedron
Lett. 2003, 44, 2991–2994; (i) Balan, D.; Adolfsson, H.
Tetrahedron Lett. 2003, 44, 2521–2524.
In conclusion, we have developed a new one-pot proce-
dure to assemble an aromatic aldehyde, a b-unsubsti-
tuted acrylate and a methide nucleophile. With the
numerous commercially available aldehydes, b-unsubsti-
tuted acrylates and activated methide nucleophiles, this
one-pot procedure of sequential Baylis–Hillman and
Michael reactions will be a useful tool for generating
highly functionalized and diversified organic molecules.
6. Aggarwal, V. K.; Mereu, A. Chem. Commun. 1999,
2311–2312.
7. Typical procedure: To a solution of benzaldehyde (1g,
1.0equiv) in THF (2mL) were added DBU (1.0equiv) and
methyl acrylate (1.1equiv) subsequently. The resulting
solution was stirred at ambient temperature for 16h.
2-Nitropropane (1.2equiv) was slowly introduced into the
reaction solution and stirred for another 3h. The reaction
mixture was concentrated to dryness and purified by silica
gel chromatography to afford the desired product as a
diastereomeric mixture.
Acknowledgements
8. All compounds were charaterized spectroscopically. Com-
pound 9 (four diastereoisomers): 1H NMR (400MHz,
CDCl3): d 7.5–7.4 (m, 2H, aromatic), 7.2–7.1 (m, 2H,
aromatic), 6.1–5.9 (m, 1H, olefinic), 5.8–5.6 (two sets, m,
1H, olefinic), 4.8–4.6 (two sets, dd+d, 1H, CH–O), 3.6–3.5
(four peaks, s, 3H, CH3–O), 3.1 (br, 1H), 2.85–2.65 (m, 1H,
CH–CO2), 2.6–1.5 (m, 8H, 4 · CH2). 13C NMR (400MHz,
CDCl3): d 174.6 (174.4), 140.3 (140.2, 140.1), 134.9 (134.8,
134.7, 134.6), 132.2 (131.9, 131.8), 128.5 (128.4, 128.2),
125.6 (125.4, 125.3, 125.1), 122.7 (122.3), 88.6 (88.5), 75.6
(74.1), 52.5 (52.4), 49.0 (48.7, 48.6), 39.8 (39.6, 37.8, 37.6),
32.4 (31.8, 31.7, 31.4), 24.9, 19.0. MS 420, 422 (M++23).
We thank G. Graczyk-Milbrandt and S. Kennedy-Gabb
of Analytical Science, GlaxoSmithKline for NMR and
MS supports.
References and notes
1. For reviews on tandem and multi-component operations:
(a) Tietze, L. F.; Beifuss, U. Angew. Chem., Int. Ed. 1993,
32, 131; (b) Wender, P. A. Frontiers in Organic Synthesis.