Catalytic, asymmetric dimerization of methylketene

Full text of DOI:10.1021/jo961721c

8
006
J . Org. Chem. 1996, 61, 8006-8007
Ca ta lytic, Asym m etr ic Dim er iza tion of
Meth ylk eten e
Sch em e 1
Michael A. Calter
Department of Chemistry, Virginia Polytechnic Institute and
State University, Blacksburg, Virginia 24061-0212
Received September 9, 1996
We hope to develop quick, inexpensive routes to
biologically important classes of molecules using cata-
lytic, asymmetric transformations. Toward this goal, we
are exploring reactions catalyzed by chrial tertiary
amines. Tertiary amines are attractive catalysts, as
these compounds are generally stable, inexpensive, and
less toxic than most transition metals. Although the
asymmetric variants of some tertiary amine-catalyzed
Sch em e 2
1
processes are known, many such reactions have not been
attempted with chiral catalysts. One example of the
latter case is the tertiary amine-catalyzed dimerization
of methylketene (Scheme 1).2,3 We report here that
cinchona alkaloids and their derivatives catalyze the
dimerization of methylketene with high enantioselectiv-
ity, yielding a product that easily transforms into a useful
synthon for polypropionate synthesis.
The tertiary amine-catalyzed dimerization of meth-
ylketene yields â-lactone 1 via a formal Claisen conden-
sation (Scheme 1). We reasoned that a chiral tertiary
amine might impose a facial bias on ammonium enolate
2
, eventually producing â-lactone 1 in an optically
enriched form. We were encouraged in this respect by
the high enantioselectivity (98% ee) realized by Wynberg
in the related cycloaddition of ketene to chloral catalyzed
by cinchona alkaloids.4 We also assumed that 1 would
be a useful polypropionate precursor. Accordingly, we
studied the dimerization of methylketene catalyzed by
various chiral, nonracemic tertiary amines.
We used Ward’s procedure to prepare a -78 °C solution
of methylketene in tetrahydrofuran (THF).5 In this
procedure, the ketene distilled away from the reaction
pot as it formed, yielding a reactant solution free of any
starting materials or byproducts. We then added the
ketene solution to 1 mol % of the amine catalyst in THF
at -78 °C (Scheme 2). The volatility and instability to
silica gel of 1 hampered the isolation of this compound,
Ta ble 1. Ca ta lysts a n d En a n tioselectivies for th e
P r od u ction of 3
entry
catalyst
quinidine
propionylquinidine
(trimethylsilyl)quinidine
quinine
% ee of 3
1
2
3
4
5
6
98 (R)
97 (R)
98 (R)
70 (S)
54 (S)
93 (S)
propionylquinine
(trimethylsilyl)quinine
that the dimerization reaction is quantitative; reaction
of the methylketene with aniline afforded approximately
0% of the corresponding amide, and reduction of purified
3
1
afforded a 70% yield of 3. While the yield for this
reaction sequence is lower than that for the existing route
4
so we added LiAlH to the reaction mixture to produce
7
primary alcohol 3 (Table 1).6
to 3, the low cost of the starting material (100 g/$18.25
for 2-bromopropionyl bromide from Aldrich) and the lack
of intermediate isolations make this route attractive. We
are currently exploring the asymmetric dimerization of
Either enantiomer of 3 is produced with high enantio-
control from the dimerization of methylketene with
readily available catalysts. Quinidine and its derivatives
afford uniformly high enantioselectivities, while quinine
and propionylquinine are notably less selective catalysts.
This trend is similar to trends observed in numerous
8
methylketene generated by alternative means and also
different methods for reduction.
Since methylketene rapidly acylates alcohols, we sought
to determine whether the active catalysts in the quinine-
or quinidine-catalyzed reactions are the free alcohols or
the propionylated alkaloids. Propionyl quinidine afforded
the same induction as quinidine (entries 1 and 2),
indicating that the ester was probably the active catalyst
in both these reactions. However, propionyl quinine was
a less enantioselective catalyst than quinine (entries 4
and 5), suggesting that the free alcohol accounted for at
least some of the product in the quinine-catalyzed reac-
tion.9
processes utilizing cinchona alkaloids as catalysts or
ligands.1
The overall yield for this reaction sequence is 20%
based on bromopropionyl bromide, regardless of the
identity of the catalyst. The following results indicate
(
(
(
(
1) Wynberg, H. Top. Stereochem. 1986, 16, 87.
2) Sauer, J . C. J . Am. Chem. Soc. 1947, 69, 2444.
3) Samtleben, R.; Pracejus, H. J . Prakt. Chem. 1972, 314, 157.
4) (a) Wynberg, H.; Staring, E. G. J . J . Am. Chem. Soc. 1982, 104,
1
66. (b) Wynberg, H.; Staring, E. G. J . J . Org. Chem. 1985, 50, 1977.
5) McCarney, C. C.; Ward, R. S. J . Chem. Soc., Perkin Trans. 1
975, 1600.
6) The enantiomeric excess of 3 was assayed by conversion to the
(
We propose the following structure for the putative
ammonium enolate derived from quinidine and meth-
1
(
corresponding Mosher ester followed by GC analysis. The configuration
of the major enantiomer produced was established by comparison of
the rotation of this compound to the rotation of 3 derived from methyl
(7) Luke, G. P.; Morris, J . J . Org. Chem. 1995, 60, 3013.
(8) Masters, A. P.; Sorensen, T. S.; Ziegler, T. J . Org. Chem. 1986,
51, 3559.
(S)-(+)-3-hydroxy-2-methylpropionate (see ref 7).
S0022-3263(96)01721-5 CCC: $12.00 © 1996 American Chemical Society

Communications
J . Org. Chem., Vol. 61, No. 23, 1996 8007
shielding a face of the enolate than in the corresponding
quinidine diastereomers. Therefore, a larger group
(TMS) is required for the quinine diastereomers to afford
a high facial bias.
The product of this reaction, 3, is a useful synthon for
polypropionate synthesis. The titanium enolate of this
compound undergoes a stereoselective aldol addition with
isobutyraldehyde.6 Derivatives of 3 are used as dipro-
pionate equivalents in the construction of oleandolide and
1
2
portions of the siphonarins and the baconipyrones.
F igu r e 1. Possible conformations of ammonium enolate 2.
Compound 1 has not been used as a chiral synthon as it
has not previously been available in an optically enriched
form. We are exploring the in situ reactions of this novel
dipropionate equivalent with nucleophiles and electro-
philes to produce tripropionates.
ylketene (Figure 1). The enolate geometry and the
quinidine conformation are based on literature prece-
dent.1
0,11
The observed facial selectivity indicates that
rotamer A is favored over rotamer B. Rotamer B might
be destabilized by the steric interaction between the
enolate methyl group and the aromatic moiety of the
catalyst. Inspection of rotamer A indicates that one face
of the enolate is relatively open toward electrophilic
attack by another molecule of methylketene, while the
other is quite hindered.
Ack n ow led gm en t. We thank the J effress Memorial
Trust for financial support of this project. We also
thank Prof. R. Gandour and Prof. J . Tanko for helpful
discussions.
The results indicate that the oxygen substituent in
quinine and its derivatives rotates further away from
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures for the preparation of 3 and all the quinine and
quinidine derivatives used as catalysts; full characterization
of quinine and quinidine derivatives (5 pages).
(
9) These results are similar to Wynberg’s results with acylated
catalysts in the addition of ketene to chloral (ref 4a).
10) There are numerous examples of nucleophiles adding to al-
(
J O961721C
doketenes opposite the larger substituent. For example, see: Dondoni,
A.; Fantin, G.; Fogagnolo, M.; Mastellari, A.; Medici, A.; Pedrini, P. J .
Org. Chem. 1984, 49, 3478.
(12) (a) Paterson, I.; Norcross, R. D.; Ward, R. A.; Romera, P.; Lister,
M. A. J . Am. Chem. Soc. 1994, 116, 11287. (b) Paterson, I.; Franklin,
A. S. Tetrahedron Lett. 1994, 35, 6925.
(11) Dijkstra, G. D. H.; Kellog, R. M.; Wynberg, H.; Svendsen, J . S.;
Marko, I.; Sharpless, K. B. J . Am. Chem. Soc. 1989, 111, 8069.