Imidazolides have been widely used in the acylation7a of
alcohols,7b amines,7c acids,7d and carbanions.7e The reaction of
ketone enolates with imidazolides favors C-acylation, and it is
a better alternative to the Claisen condensation to synthesize
â-keto esters and 1,3-diketones.8 As we expected, treatment of
the sodium enolate of 1-tetralone in DME at -78 °C with allyl
1H-imidazole-1-carboxylate 1a gave only C-acylation product
3 (eq 1). It is a general notion that reaction of ketone enolates
with “hard” electrophiles favor O-acylation while “soft” elec-
trophiles favor C-acylation.6 We postulated that the complex
of the imidazolide with an azaphilic Lewis acid, such as boron
trifluoride, is a “harder” electrophile than the imidazolide itself
and it should favor O-acylation.9 To our delight, the experi-
mental result supported our proposal, and in contrast to eq 1,
the reaction mediated with BF3‚Et2O cleanly generated only allyl
enol carbonate 4a in 85% yield (eq 2).
TABLE 1. Formation of Allyl Enol Carbonates from Various
Substituted Allyl 1H-Imidazole-1-carboxylates
To investigate the generality of the reaction, we prepared
several substituted allyl 1H-imidazole-1-carboxylates 1 by
simply mixing 1,1′-carbonyldiimidazole with the corresponding
allylic alcohols in a cosolvent of THF and CH2Cl2 at 0 °C for
several hours and then isolated the product by silica gel column
chromatography. The yields are normally very good (Table 1).
As a demonstration, the sodium enolate of 1-tetralone or
2-methyl-1-tetralone was treated with a variety of the BF3
complexes of differently substituted allyl derivatives 1, respec-
tively, and consistently, good yields were obtained in every case.
It is remarkable that, through this procedure, enol carbonates
containing amino groups could be produced in the same manner
(Table 1, entry 10).
The reaction of various sodium enolates with the BF3 complex
of 1 exclusively gave enol carbonates in high yields (Table 2,
entries 7, 9, and 11), while the ones without BF3 gave
exclusively â-keto esters in Table 2, entries 1 and 6, or a mixture
of both in entry 8, possibly because the steric effect of the methyl
group disfavored the C-acylation. There are also some cases in
which the nature of the enolate is such that the regioselectivity
toward the O-acylation is not perfect even with the BF3-activated
electrophile. For instance, the reaction of the sodium enolate
of 1-indanone with the 1d-BF3 complex in DME at -78 °C
yielded a mixture of enol carbonate 8 and â-keto ester 7 in a
3:1 ratio (Table 2, entry 2). By switching to the less coordinated
potassium enolate, generated by use of potassium tert-butoxide
as base, the ratio improved to 17:1 (Table 2, entry 3). In the
presence of 18-C-6, the enolate has even weaker coordination
with the potassium countercation, and in this case, no 7 was
observed (Table 2, entry 4). Since the reactions between enolates
and 1-BF3 complex are fast even at -78 °C, either the
thermodynamic (such as 13 in entry 9) or the kinetic product
(such as 15 in entry 10) can be obtained with excellent selectivity
a Isolated yields of 1 from the reaction of 1,1′-carbonyldiimidazole and
the corresponding allyl alcohol.
by quenching the corresponding enolates generated using
different conditions. Substrates 16 and 17 have been used in
the formal synthesis of (S)-oxybutynin, and they can be
alternatively prepared by this method (Table 2, entry 11).1d
In summary, allyl 1H-imidazole-1-carboxylates are readily
accessible from 1,1′-carbonyldiimidazole and allylic alcohols.
They are normally stable and easy to handle. By tuning their
reactivity with BF3 in the reaction with ketone enolates, we have
developed a convenient procedure for the synthesis of a broad
scope of allyl enol carbonates in high yields and regioselectivity.
The study of Pd-catalyzed asymmetric decarboxylative allylic
alkylation reactions of substituted allyl enol carbonates and their
application in organic synthesis is ongoing.
Experimental Section
(7) (a) For review concerning the application of imidazolides as acylating
agents, see: Staab, H. A. Angew. Chem., Int. Ed. Engl. 1962, 7, 350. (b)
Bertolini, G.; Pavich, G.; Vergani, B. J. Org. Chem. 1998, 63, 6031. (c)
Vatele, J.-M. Tetrahedron 2004, 60, 4251. (d) Kitagawa, T.; Kuroda, H.;
Sasaki, H. Chem. Pharm. Bull. 1987, 35, 1262. (e) Werner, T.; Barrett, A.
G. M. J. Org. Chem. 2006, 71, 4302.
(8) (a) Moyer, M. P.; Feldman, P. L.; Rapoport, H. J. Org. Chem. 1985,
50, 5223. (b) Shone, R. L.; Deason, J. R.; Miyano, M. J. Org. Chem. 1986,
51, 268. (c) Tanaka, T.; Okamura, N.; Bannai, K.; Hazato, A.; Sugiura, S.;
Tomimori, K.; Manabe, K. i.; Kurozumi, S. Tetrahedron 1986, 42, 6747.
(9) For reference concerning the much better reactivity of acylimidazo-
lium salts than the corresponding imidazolides, see: (a) Anders, E.; Will,
W. Tetrahedron Lett. 1978, 41, 3911. (b) Oakenfull, D. G.; Salvesen, K.;
Jencks, W. P. J. Am. Chem. Soc. 1971, 93, 188. (c) Grzyb, J. A.; Shen, M.;
Yoshina-Ishii, C.; Chi, W.; Brown, R. S.; Batey, R. A. Tetrahedron 2005,
61, 7153.
General Procedures. Procedure A: A clean oven-dried 50 mL
flask was charged 460 mg of sodium bis(trimethylsilyl)amide (2.4
mmol) under nitrogen. The flask was cooled to -78 °C in a dry
ice/acetone bath, and 5 mL of 1,2-dimethoxyethane (DME) was
added through a syringe. The flask was taken out from the bath to
allow it to warm up until the solid completely dissolved. Then it
was cooled to -78 °C. To the solution was added 317 mg of
1-indanone (2.4 mmol) in 2 mL of DME. The solution was stirred
for 30 min before it was transferred into the solution of 384.4 mg
of 1d (2 mmol). After 10 min, the cooling bath was removed and
the reaction was allowed to warm to room temperature and stirred
at room temperature for 1 h. One portion of 10 mL of saturated
aqueous ammonium chloride was poured into the reaction mixture
J. Org. Chem, Vol. 72, No. 24, 2007 9373