J . Org. Chem. 1997, 62, 4327-4329
4327
Lew is Acid -P r om oted Cou p lin g Rea ction s of Acid Ch lor id es w ith
Or ga n oa lu m in u m a n d Or ga n ozin c Rea gen ts
Mitsuhiro Arisawa, Yasuhiro Torisawa, Michiaki Kawahara, Masamichi Yamanaka,
Atsushi Nishida, and Masako Nakagawa*
Faculty of Pharmaceutical Sciences, Chiba University, 1-33 Yayoi-cho, Inage-ku, Chiba 263, J apan
Received February 11, 1997X
An efficient synthesis of R,â-unsaturated ketones by the reaction of acid chlorides with trialkyl-
aluminum (1/3 mole equiv) in the presence of AlCl3 (1 mol equiv) is described. Dialkylzincs were
also useful and are easier to prepare than trialkylaluminum. Reaction of RCOCl with R′AlCl2 or
R′2AlCl gave R′COR, without AlCl3, in high yield.
Sch em e 1
The coupling reactions of alkylmetal reagents and acid
halides provide useful method for synthesizing
a
ketones.1-3 During our study of the Diels-Alder reac-
tion,4 we became interested in reacting acid chlorides
with organoaluminum reagents to give ketones in a single
operation. Inspired by some earlier reports,2,3 we devel-
oped a practical method for producing (E)-pent-3-en-2-
one (1) (Scheme 1), a useful precursor for some silyloxy
dienes, by reacting crotonyl chloride and triethylalumi-
num.5 In this report, we describe this reaction in detail
and its extension to an organozinc reagent.
Sch em e 2
When a CH2Cl2 solution of trimethylaluminum (0.4
equiv) was added to a solution of crotonyl chloride and
aluminum chloride (1.0 equiv) in CH2Cl2, alkylation
proceeded to give (E)-pent-3-en-2-one (1) quantitatively
(Table 1). The presence of aluminum chloride was
essential; without aluminum chloride, only 9% of 1 was
obtained. Although aluminum fluoride and gallium
chloride were less effective, methylaluminum chloride
and dimethylaluminum chloride were effective for the
alkylation of crotonyl chloride.
aluminum reagent, such as trioctylaluminum, also worked
without any problems.
The general procedure, which has been reported previ-
ously,5 was generally reproducible for the preparation of
a few grams of 1. However, we encountered a serious
problem when we tried to prepare a large quantity (>10
g) of 1. Specifically, the product was contaminated by a
varying amount of 3-chloropentan-2-one, and it was
difficult to separate these compounds because of their
close boiling points. The amount of contaminating chloro
ketone depended on the workup prodedures (1:chloro
ketone ) 1:1-10:1), such as the washing time using
aqueous base and the time necessary for removal of
CH2Cl2 under atmospheric pressure. Finally, we estab-
lished two procedures for the large-scale preparation of
1. One procedure consists of the elimination of hydrogen
chloride from the chloro ketone to 1, basically as de-
scribed in ref 6. The product obtained using this protocol
sometimes contained a trace amount of â,γ-unsaturated
ketone.
The reaction of phenylpropionyl chloride with tri-
methylaluminum was interesting from a mechanistic
point of view. When a solution of aluminum chloride was
added to a mixture of the acid chloride and trimethyl-
aluminum, the desired propiophenone was obtained in
91% yield (Scheme 2). However, the addition of tri-
methylaluminum to a mixture of aluminum chloride and
the acid chloride under the same conditions gave 1-in-
danone quantitatively. These results suggested that the
reaction did not proceed via the simple acylinium ion
complex (A in Figure 1), but rather proceeded via the acyl
chloride-AlR′3 complex (B) activated by the AlClXY
species (X, Y ) Cl or alkyl) as shown in Figure 1. A bulky
Since the addition of HCl to 1 appeared to occur when
CH2Cl2 was removed by distillation under atmospheric
pressure, the second procedure included the rapid distil-
lation of a mixture of CH2Cl2 and 1 under reduced
pressure at less than 50 °C using a cold trap (-78 °C).
The solvents were then distilled in the presence of
quinoline. Purification of the residue by distillation
under vacuum gave on the order of 30 g of pure 1.
Regarding the preparation of organometallic reagents,
dialkylzinc reagents are much easier to prepare and
sometimes more stable than trialkylaluminums.1,7 Thus,
we next focused our attention on the use of dimethylzinc
X Abstract published in Advance ACS Abstracts, J une 1, 1997.
* To whom correspondence should be addressed. Tel.: 81-43-290-
2908. Fax: 81-43-255-1574. E-mail: nakagawa@p.chiba-u.ac.jp.
(1) See as general references: (a) Tamao, K. In Comprehensive
Organic Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press:
New York, 1991; Vol. 3, pp 463-464. (b) Larock, R. C. Comprehensive
Organic Transformations; VCH Publishers: New York, 1989; pp 686-
691. (c) Posner, G. A. Org. React. 1975, 22, 253-400. (d) Shirley, D. A.
Org. React. 1954, 8, 28-58.
(2) Tolstikov, G. A.; Valitov, F. K.; Kuchin, A. V. J . Gen. Chem.
(USSR) 1982, 51, 1359-1361.
(3) (a) Wakamatsu, K.; Okuda, Y.; Oshima, K.; Nozaki, H. Bull.
Chem. Soc. J pn. 1985, 58, 2425-2426. (b) Takai, K.; Oshima, K.;
Nozaki, H. Ibid. 1987, 54, 1281-1282.
(4) Nagata, T.; Koide, Y.; Nara, K.; Itoh, E.; Arisawa, M.; Naruto,
S.; Torisawa, Y.; Hino T.; Nakagawa, M. Chem. Pharm. Bull. 1996,
44, 451-453.
(5) Arisawa, M.; Torisawa, Y.; Nakagawa, M. Synthesis 1995, 1371-
1372.
(6) Odom, H. C.; Pinder, A. R. Organic Syntheses; Wiley: New York,
1988; Collect. Vol. VI, pp 883-886.
S0022-3263(97)00244-2 CCC: $14.00 © 1997 American Chemical Society