Catalytic, Asymmetric Preparation of Ketene Dimers from Acid Chlorides

Article abstract of DOI:10.1021/ol0359517

(Matrix presented) The cinchona alkaloid-catalyzed dimerization of monosubstituted ketenes generated in situ from the reaction of acid chlorides and diisopropylethylamine yields ketene dimers in high yields and enantioselectivities. This reaction tolerates sterically demanding and functionally diverse substituents. Kinetic studies suggest that the rate-determining step for the reaction is the deprotonation of the acid chloride by the tertiary amine to form ketene and that the stereochemistry- forming step is addition of an ammonium enolate with ketene.

Full text of DOI:10.1021/ol0359517

ORGANIC
LETTERS
2003
Vol. 5, No. 24
4745-4748
Catalytic, Asymmetric Preparation of
Ketene Dimers from Acid Chlorides
Michael A. Calter,* Robert K. Orr, and Wei Song
Department of Chemistry, UniVersity of Rochester, Rochester, New York 14627-0216
calter@chem.rochester.edu
Received October 6, 2003
ABSTRACT
The cinchona alkaloid-catalyzed dimerization of monosubstituted ketenes generated in situ from the reaction of acid chlorides and
diisopropylethylamine yields ketene dimers in high yields and enantioselectivities. This reaction tolerates sterically demanding and functionally
diverse substituents. Kinetic studies suggest that the rate-determining step for the reaction is the deprotonation of the acid chloride by the
tertiary amine to form ketene and that the stereochemistry-forming step is addition of an ammonium enolate with ketene.
Ketenes have served as key intermediates in a number of
important reactions, particularly those involving asymmetric
catalysis.1 However, the general instability of ketenes has
complicated methods for their generation. Fortunately,
several groups recently reported catalytic, asymmetric reac-
tions of ketenes generated in situ from the reaction of acid
halides with tertiary amines.2 We previously reported that
the cinchona alkaloid-catalyzed dimerization of pyrolytically
generated methylketene efficiently and enantioselectively
afforded a highly useful intermediate for the synthesis of
polypropionates.3 We now report the first examples of
catalytic, asymmetric dimerization of in situ generated
ketenes, along with the dramatically expanded scope possible
with the new conditions. We also show that this reaction is
mechanistically distinct from other, nucleophile-catalyzed
reactions of in situ generated ketenes.
of ketenes from acid chlorides and ketene dimerization.
Dimerization of pyrolytically formed methylketene required
nucleophilic catalysis by cinchona alkaloid derivatives. This
requirement necessitated the choice of a stoichiometric base
that would not catalyze the formation of racemic dimer.4
Several groups have described the use of cinchona alkaloid
catalysts and non-nucleophilic, stoichiometric amines to limit
stoichiometric amine-catalyzed reactions. In one of the
examples most relevant to the desired reaction, Lectka and
co-workers reported that acid chlorides condense with
reactive imines at -78 °C using catalytic benzoylquinine
(BzQN) and stoichiometric proton sponge (PS) in toluene
(Scheme 1).2d In related work, Romo et al. used quinidine
(QD) and diisopropylethylamine (Hu¨nig’s base, HB) in
toluene at -20 °C for the addition of acetyl chloride to
reactive aldehydes.2c However, neither of these precedents
offered compelling evidence for the intermediacy of a ketene.
Our design of conditions for the target reaction started with
precedents for the two basic steps: base-promoted formation
Based on the precedents of Scheme 1, we assayed
conditions for the in situ generation and dimerization of
methylketene, starting with propionyl chloride (Scheme 2,
Table 1). For the purposes of determining yield and enan-
tiomeric purity, we immediately converted volatile and
unstable methylketene dimer 1 into â-ketoamide 2.5 Treat-
(1) For a recent review of asymmetric reactions involving ketenes, see:
Orr, R. K.; Calter, M. A. Tetrahedron 2003, 59, 3545-3565.
(2) (a) Nelson, S. G.; Peelen, T. J.; Wan, Z. J. Am. Chem. Soc. 1999,
121, 9742-9743. (b) Yoon, T. P.; MacMillan, D. W. C. J. Am. Chem. Soc.
2001, 123, 2911-2912. (c) Tennyson, R. L.; Romo, D. J. Org. Chem. 2000,
65, 7248-7252. (d) Taggi, A. E.; Hafez, A. M.; Wack, H.; Young, B.;
Drury, W. J., III; Lectka, T. J. Am. Chem. Soc. 2000, 122, 7831-7832.
(3) Calter, M. A.; Liao, W. J. Am. Chem. Soc. 2002, 124, 13127-13129
and references therein.
(4) Sauer, J. C. J. Am. Chem. Soc. 1947, 69, 2444-2448.
10.1021/ol0359517 CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/23/2003

Table 1. Yields and Selectivities for the Formation of 2a
Scheme 1
entry catalystb solvent base % yield of 2 % ee 2 (config)
1
2
3
4
5
6
7
8
9
PhMe
PhMe
PhMe
PhMe
CH2Cl2 HB
CH2Cl2 HB
CH2Cl2 HB
CH2Cl2 HB
CH2Cl2 HB
CH2Cl2 HB
CH2Cl2 HB
PS
HB
PS
0
0
0
BzQN
TMSQN
TMSQN
QN
HB
19
79
56
64
65
79
72
71
71
93 (R)
94 (R)
70 (R)
80 (R)
69 (R)
97 (S)
94 (R)
96 (R)
97 (R)
BzQN
PropQN
TMSQD
TBSQN
TMSQNc
10
11
12
TMSQNc,d CH2Cl2 HB
a All reactions run starting with 2 mmol of propionyl chloride at a
concentration of 0.1 M, unless otherwise noted. b PropQN ) propionyl
quinine. c Filtered through silica gel prior to formation of 2. d Run starting
with 115 mmol of propionyl chloride in 460 mL of CH2Cl2 (0.25 M).
pure form and reduced yield by rapid filtration of the
unpurified reaction mixture through a small plug of silica
gel, a finding that has importance for other reactions of 1
(entry 11). Finally, increasing the scale and concentration
did not effect the yield or enantioselectivity of the reaction
(entry 12).
ment with stoichiometric PS or HB in toluene did not lead
to product in the absence of cinchona alkaloid catalyst
Scheme 2
Table 2. Yields and Selectivities for the Formation of 4a-e
compd
R
time A (h) time B (h) % yield % ee
4a
4b
4c
4d
4e
Et
iPra
tBua
CH2OTIPS
CH2CO2Meb
6
24
96
3
2
24
24
2
82
65
58
88
64
92
96
92
91
92
3
2
a Reaction performed at 0.5 M concentration of acid chloride. b Dimer
filtered through silica gel before opening.
(entries 1 and 2). Lectka’s conditions from Scheme 1 also
failed to produce significant quantities of 2 (entry 3).
However, these workers have shown that pathways not
involving ketene intermediates predominate under these
conditions.6 Conditions similar to those of Romo did afford
minor amounts of product (entry 4), but the key to high levels
of conversion was the use of CH2Cl2 as solvent (entry 5).
We had previously demonstrated that trimethylsilylquinine
(TMSQN) was superior to the parent alkaloid or acyl
derivatives in the dimerization of preformed methylketene.
The same held true in the present reaction, as the nonsilylated
alkaloids afforded lower selectivities and conversions (entries
7 and 8). Conveniently, the use of trimethylsilylquinidine
(TMSQD) gave high enantioselectivity for the opposite
enantiomer (entry 9). One could also isolate 1 in relatively
The conditions developed for 1 also allowed the prepara-
tion of dimers 3a-e, bearing a variety of more highly
functionalized side chains not compatible with pyrolytic
methods for ketene generation (Scheme 3, Table 2). Again,
the dimers were converted without isolation into the corre-
Scheme 3
(5) Previous experiments demonstrated that the conversion of 1 into 2
proceeds in high yield and without epimerization. See: Calter, M. A.; Guo,
X. J. Org. Chem. 1998, 63, 5308-5309.
(6) Taggi, A. E.; Hafez, A. M.; Wack, H.; Young, B.; Ferraris, D.; Lectka,
T. J. Am. Chem. Soc. 2002, 124, 6626-6635.
4746
Org. Lett., Vol. 5, No. 24, 2003

sponding â-ketoamides 4a-e. The reaction tolerated increas-
ing steric bulk, although adequate conversion to the isopropyl
and tert-butylketene dimers required an initial concentration
of 0.5 M for the acid chloride and longer times for the
opening reaction. Substrates bearing convenient functional
groups for further elaboration, such as â-siloxy or ester
groups, afforded dimers in reasonable yields and high
enantioselectivities.
mediate between acid chloride and BzQN was the rate-
determining step.6
We next sought to determine whether the nature stereo-
chemistry-determining step of the dimerization depended on
whether the methylketene was pregenerated or generated in
situ. All evidence from our work and the work of others
indicated that the stereochemistry-determining step involved
the reaction of an ammonium enolate, such as 5, with an
electrophile (Scheme 5).8 In the case of the dimerization of
We next determined the rate- and stereochemistry-
1
determining step of the reaction. A simple H NMR study
showed that the rate of generation of 1 from propionyl
chloride depended linearly on the concentration of propionyl
chloride and HB, but not on catalyst concentration (Figure
1).7 These results were consistent with ketene formation
Scheme 5
pregenerated methylketene (Scheme 6), the electrophile must
be another molecule of methylketene (path A, Scheme 5).9
However, the use of in situ generated methylketene opens
the possibility of a reaction between 5 and propionyl chloride
(path B).
Figure 1. Initial rates for the formation of 1 at various concentra-
tions of propionyl chloride, HB, and TBSQN. Standard conditions
refer to starting concentrations of 0.1 M for both propionyl chloride
and HB and 0.005 M for TBSQN. The reactions were monitored
by 1H NMR at 25 °C in CD2Cl2, and the amount of 1 was quantified
by comparison of the integration of the signal at 4.75 ppm with
that of the signal at 5.82 ppm for an internal standard, 2,5-
dimethylfuran.
Scheme 6
being rate-determining for the in situ dimerization reaction
and the rate-determining step for ketene formation being
deprotonation of acid chloride by HB (Scheme 4). The large
In an attempt to distinguish between paths A and B for
the dimerization of in-situ-generated methylketene, we
compared the optical activity of 1 produced under these
conditions with that produced from thermolytically generated
methylketene (Schemes 2 and 6, Table 3). Using several
catalysts with a range of enantioselectivies, the optical
activities were identical within experimental error. This result
suggested a common stereochemistry-determining step for
both reactions, and therefore, that the dimerization of in situ-
generated methylketene proceeded by path A. This conclu-
sion implies that methylketene was significantly more
electrophilic toward 5 than propionyl chloride, as the results
from the rate study indicated that concentration of meth-
ylketene was generally much lower than that of propionyl
chloride.
Scheme 4
kinetic isotope effect, 5.1, observed starting with propionyl
chloride-2,2-d2 supported these conclusions. This mechanism
differed dramatically from that for the Lectka example in
Scheme 1, in which formation of an acylammonium inter-
(7) The kinetics experiments were conducted using TBSQN as catalyst,
as this compound demonstrated optimal stability under the reaction
conditions.
(8) Pracejus, H. Ma¨tje, H. J. Prakt. Chem. 1964, 24, 195-205.
(9) Samtleben, R.; Pracejus, H. J. Prakt. Chem. 1972, 314, 157-169.
Org. Lett., Vol. 5, No. 24, 2003
4747

differs dramatically from that of previously studied, nucleo-
phile-catalyzed reactions of in situ generated ketenes. The
discovery of these conditions should impact the use of in
situ generated ketenes in other catalytic, asymmetric pro-
cesses.
Table 3. Yields and Selectivities for the Formation of 2a
reactiona
catalyst
% ee of 2
Scheme 2
Scheme 6
Scheme 2
Scheme 6
Scheme 2
Scheme 6
TMSQN
TMSQN
BzQN
BzQN
PropQN
PropQN
96
97
80
81
69
69
Acknowledgment. We acknowledge the NIH for support
and Prof. Joseph Dinnocenzo for helpful conversations.
a Results for Scheme 2 from Table 1, entries 7, 8, and 11.
Supporting Information Available: Complete experi-
mental procedures and characterization data for compounds
2 and 4a-e. This material is available free of charge via the
Internet at http://pubs.acs.org.
In conclusion, we have developed conditions for the in
situ generation and dimerization of ketenes from acid halides.
These conditions allowed a fundamental expansion in the
scope of the dimerization. The mechanism of the reaction
OL0359517
4748
Org. Lett., Vol. 5, No. 24, 2003