Homogeneous catalysis. Mechanism of the catalysed Mukaiyama cross-aldol reaction

Article abstract of DOI:10.1039/a705801b

It is shown that <2 + 2> addition intermediates can form reversibly in the catalysed Mukaiyama reaction; these serve to prevent formation of the catalyst Me3Si(1+) and the reactivities of the <2 + 2> intermediates can affect the enantioselectivity of the

Full text of DOI:10.1039/a705801b

View Article Online / Journal Homepage / Table of Contents for this issue
Homogeneous catalysis. Mechanisms of the catalysed Mukaiyama cross-aldol
reaction
William W. Ellis and B. Bosnich*
Department of Chemistry, The University of Chicago, 5735 South Ellis Avenue, Chicago, Illinois 60637, USA
H
It is shown that [2 + 2] addition intermediates can form
OSiMe3
+
O
reversibly in the catalysed Mukaiyama reaction; these serve
to prevent formation of the catalyst Me3Si+ and the
reactivities of the [2 + 2] intermediates can affect the
enantioselectivity of the reaction.
OMe
Ph
3
C3F7
O
C6D6, 20 °C
The Mukaiyama cross-aldol reaction is a versatile carbon–
carbon bond-forming reaction which is catalysed by a variety of
Lewis acids.1 It is generally assumed to proceed by way of a
cyclic intermediate 1 which allows for transfer of the silyl group
(Scheme 1). For certain Lewis acids, however, intramolecular
silicon transfer does not occur, rather the trimethylsilyl group is
captured by the carbonyl substrate2,3 and catalysis occurs by
means of the powerful Me3Si+ catalyst.2 Despite the possible
widespread intervention of the (achiral) Me3Si+ catalytic path,
there are a number of reports of chiral Lewis acid catalysts
which give excellent enantiomeric excesses (ees),4 indicating
that for these catalysts, the unwanted Me3Si+ path is suppressed.
The question then arises as to which type of Lewis acids will
allow for one or other of the possible catalytic paths. It is
probable that if the Lewis acid–oxygen bond of the aldolate 1 is
weak and the bound oxygen atom is a good nucleophile, the rate
of the silyl transfer step will be enhanced and consequently the
capture of the Me3Si+ group by external nucleophiles will be
suppressed. The direct silyl transfer path as illustrated in
Scheme 1, however, may not be the only component of the
mechanism because it is possible that an oxetane could form by
capture of the carbenium ion of the intermediate. It is known
that oxetanes are formed by Lewis acid-catalysed coupling of
aldehydes and dialkyl ketene acetals.5 The formation of the
oxetane would be the equivalent of the Mukaiyama reaction, as
an acidic workup would provide the desired aldol product. We
report here on the detection of oxetanes in the Mukaiyama
reaction using lanthanide and zinc complexes, both of which
were expected to form aldolate intermediates with weak metal–
oxygen bonds.
O
Eu
2
3
MeO
O
Me3SiO
O
OSiMe3
OMe
+
Ph
Scheme 2
Ph
4
5
the equilibrium illustrated in Scheme 2 obtains (K ≈ 3 at 20 °C).
Addition of 1 m pyridine quenches catalysis, but the addition of
the hindered base, 2,6-di-tert-butyl-4-methylpyridine, does not
affect the rate of catalysis. This latter observation indicates that
the reaction in Scheme 2 is neither proton-initiated nor proton-
catalysed. Under these conditions—4 mol% catalyst, 1 m each
of substrates, benzene solution, 20 °C—the Mukaiyama product
begins to appear after several hours and, after seven days, all of
the substrates and intermediates are converted to this product.
The Mukaiyama product is irreversibly formed because a 15%
ee (S)6 was found using the (+)-hfc catalyst, and the ee of the
product did not change in the presence of the catalyst after two
weeks. It was found that, after equilibration of the two oxetanes,
their hydrolysed product had 0% ee using the (+)-hfc catalyst
but a 5% ee (S) was observed for the hydrolysed products
derived from the kinetically formed oxetanes (ca. 10%
conversion to oxetanes).‡ This observation also supports the
view that equilibration between substrates and oxetanes exists
(Scheme 2).
The results are consistent with the mechanism outlined in
Scheme 3 (Ln = 2). In this case, the rate of production of the
oxetanes by carbenium ion capture and the reverse reaction are
faster than silyl transfer to produce the Mukaiyama product 7.
As might be expected, these relative rates vary according to the
nature of the substrates and the catalyst. This is illustrated by a
In benzene solution at 20 °C, the catalyst [Eu(hfc)3] 2 (4
mol%) promotes the reaction between benzaldehyde and the
ketene acetal 3 to give initially the two oxetane isomers 4 and 5
(Scheme 2). We have not been able to isolate these oxetanes in
pure form but they have been characterized by their 1H and 13
C
NMR spectra, although the identity of the isomers 4 and 5 was
not established.† The initial (kinetic) ratio of the isomers is
48:52 which slowly reaches an equilibrium ratio of 38:62. At
1 m concentration of each of the substrates, a maximum yield of
56% for the two oxetanes is obtained after 1 h in benzene
solution at 20 °C using 4 mol% catalyst. Isomer equilibrium
occurs over 2 h. The maximum yield of oxetanes (4 and 5)
decreases upon dilution of the reaction solution, indicating that
–
Ln
O
Ln
O
OSiMe3
+
Ln
+
4
5
+
3
–Ln
–Ln
Ph
H
Ph
OMe
6
Ln
Me3
Me3Si
Si
–
LA
R1
H
O
O
Me3SiO
R1
O
OSiMe3
XR2
O
O
LA
+
O
+
XR2
XR2
R1
Ph
OMe
X = O, S
LA
7
Scheme 3
1
Scheme 1
Chem. Commun., 1998
193

View Article Online
H
R
Me3SiO
Eu(hfc)3
of the diastereomeric species in the mechanism. Under these
circumstances, there would not be a single chirality-controlling
step. The observations presented here may help in the design of
effective enantioselective catalysts for the Mukaiyama reac-
tion.
OSiMe3
OSiMe3
+
+
O
Me3SiO
O
Me3SiO
O
This work was supported by grants from the NIH.
OSiMe3
OSiMe3
OSiMe3
Footnotes and References
R
R
OSiMe3
trans
* E-mail: bos5@midway.uchicago.edu
cis
† Selected spectroscopic data for the major and minor oxetanes (at
equilibrium): dH (400 MHz, C6D6) (major isomer) 7.3–7.1 (m, 5 H), 4.95 (s,
1 H), 3.26 (s, 3 H), 1.27 (s, 3 H), 0.76 (s, 3 H), 0.29 (s, 9 H); (minor isomer)
7.3–7.1 (m, 5 H), 4.99 (s, 1 H), 3.34 (s, 3 H), 1.25 (s, 3 H), 0.76 (s, 3 H), 0.23
(s, 9 H); dC (125 MHz, CDCl3) (major isomer) 139.0 (s), 127.7 (d), 127.2
(d), 125.1 (d), 116.2 (s), 82.6 (d), 49.4 (s), 48.8 (q), 23.5 (q), 21.1 (q), 1.05
(q); (minor isomer) 138.5 (s), 127.7 (d), 127.0, 125.6 (d), 115.7 (s), 81.8 (q),
50.7 (s), 49.0 (q), 19.0 (q), 18.1 (q), 1.06 (q).
Eu(hfc)3
Me3SiO
R
O
Me3SiO
O
+
OSiMe3
OSiMe3
erythro
R
OSiMe3
OSiMe3
threo
‡ Mild acid hydrolysis of the mixture of oxetanes leads to the formation of
the b-hydroxy ester. Enantiomeric excesses of the hydrooxy ester were
determined using Eu(tfc)3 as a chiral shift reagent.
R = Ph, Me
Scheme 4
§ The cis and trans oxetanes where R = Ph and Me were identified by
conversion of mixtures of these compounds to mixtures of their respective
threo and erythro 2,3-dihydroxyalkanoic acids according to the reported
procedure (ref. 7). In both cases, where R = Ph or Me, it is noted that
hydrolysis of the cis and trans oxetanes leads to formation of the threo and
erythro products, respectively, which were identified by comparison with
spectroscopic data from the literature (ref. 8).
Selected spectroscopic data for the oxetanes: dH (400 MHz, C6D6) (R =
Ph, threo) 7.47 (m, 2 H), 7.2–7.0 (m, 3 H), 5.18 (d, J 6.2, 1 H), 4.69 (d, J
6.2, 1 H), 0.29 (s, 9 H), 0.27 (s, 9 H), 20.22 (s, 9 H); (R = Ph, erythro) 7.43
(m, 2 H), 7.2–7.0 (m, 3 H), 4.98 (d, J 4.9, 1 H), 4.35 (d, J 4.8, 1 H), 0.35 (s,
9 H), 0.26 (s, 9 H), 0.01 (s, 9 H); (R = Me, threo) 4.42 (d, J 6.3, 1 H), 4.28
(m, 1 H), 1.25 (d, J 6.4, 3 H); (R = Me, erythro) 4.10 (m, 1 H), 4.02 (d, J
4.8, 1 H), 1.21 (d, J 6.2, 3 H). The OSi(CH3)3 resonances for R = Me were
left unassigned due to the complexity of this region in the 1H NMR spectrum
of the reaction mixture.
comparison of the reactions shown in Scheme 4, where it is
found that in catalysis of benzaldehyde, a cis:trans ratio of
80:20 is observed for the Mukaiyama product. For 1 mol%
catalyst at 20 °C in benzene solution, the benzaldehyde reaction
gives a cis:trans ratio of 75:25 at 40% conversion to the two
oxetanes and 55:45 after 60% conversion.§ This change in
isomer ratio occurs before significant amounts of the Mu-
kaiyama products are formed, indicating that equilibration
between the substrates and oxetanes occurs. Final equilibrium
may not be established under these conditions, however,
because a constant cis:trans ratio was not observed before the
Mukaiyama product began to appear. For the analogous
reaction with acetaldehyde (Scheme 4), under the same
conditions, the rates of formation of the oxetanes and the
Mukaiyama product are comparable. Similarly, the [Zn(fa-
cac)2·2H2O] (facac = hexafluoroacetylacetonate) complex
catalyses the coupling between benzaldehyde and 3 but the
equilibrium 4 " 5 is not established before significant amounts
of the Mukaiyama product is formed (Scheme 2). With this zinc
catalyst, no [2 + 2] addition products are observed in the
coupling of benzaldehyde and CH2NC(OSiMe3)SBut indicating
that catalysis proceeds either wholly by direct silyl transfer or,
if [2 + 2] products are formed, their concentrations are very low.
We note that neither the Eu nor the Zn catalysts lead to coupling
of ketones with silyl ketene acetals nor coupling of silyl enol
ethers with aldehydes or ketones.
The discovery of the [2 + 2] addition path for the catalysed
Mukaiyama is significant in a number of respects. For cases
where silyl transfer is slow in the intermediate 6 (Scheme 3), the
more rapid oxetane formation will reduce the life-time of this
intermediate. Consequently, the probability of Me3Si+ capture
in 6 by external nucleophiles will be reduced. The Me3Si+ group
is stable in the oxetanes. As was noted earlier, the mechanism
illustrated in Scheme 3 is more likely to occur when the Lewis
acid–alkoxide bond of the intermediate 6 is weak. Intermediates
having strong alkoxide–Lewis acid bonds are unlikely to lead to
oxetane formation or to silyl transfer. This appears to be the case
for many catalysts which only serve as initiators for the
production of Me3Si+ catalyst.2
1 S. Murata, M. Suzuki and R. Noyori, J. Am. Chem. Soc., 1980, 102, 3248;
C. Mukai, S. Hashizome, K. Nagami and M. Hanaoka, Chem. Pharm.
Bull., 1990, 38, 1509; H. Sukurai, K. Sasaki and A. Hosami, Bull. Chem.
Soc. Jpn., 1983, 56, 3195; N. Iwasawa and T. Mukaiyama, Chem. Lett.,
1987, 463; T. Mukaiyama, S. Kobayashi, M. Tamura and Y. Sagawa,
Chem. Lett., 1987, 491; T. Mukaiyama, S. Kobayashi and M. Murakami,
Chem. Lett., 1985, 447; T. Mukaiyama, S. Kobayashi and M. Murakami,
Chem. Lett., 1984, 1759; S. Kobayashi, M. Murakami and T. Mukaiyama,
Chem. Lett., 1985, 1535; M. T. Reetz and A. E. Vougioukas, Tetrahedron
Lett., 1987, 28, 793; S. Sato, I. Matsuda and Y. Izumi, Tetrahedron Lett.,
1987, 28, 6657; S. Sato, I. Matsuda and Y. Izumi, Tetrahedron Lett.,
1986, 27, 5517; T. Mukaiyama, T. Soga and H. Takenoshita, Chem. Lett.,
1989, 1273; S. Kobayashi, I. Hachiya and T. Takahori, Synthesis, 1993,
371; S. Kobayashi and I. Hachiya, Tetrahedron Lett., 1992, 33, 1625; S.
Kobayahi, Chem. Lett., 1991, 2187; S. Kobayashi, I. Hachiya, H. Ishitani
and M. Araki, Synlett, 1993, 472; T. Bach, D. N. A. Fox and M. T. Reetz,
J. Chem. Soc., Chem. Commun., 1992, 1634; R. Hara and T. Mukaiyama,
Chem. Lett., 1989, 1909; T. Mukaiyama and R. Hara, Chem. Lett., 1989,
1171.
2 T. K. Hollis and B. Bosnich, J. Am. Chem. Soc., 1995, 117, 4570.
3 E. M. Carreira and R. A. Singer, Tetrahedron Lett., 1994, 35, 4323.
4 E. M. Carreira, R. A. Singer and W. Lee, J. Am. Chem. Soc., 1994, 116,
8837; R. A. Singer and E. M. Carreira, J. Am. Chem. Soc., 1995, 117,
12 360; E. R. Parmee, O. Tempkin and S. Masamune, J. Am. Chem. Soc.,
1991, 113, 9365.
5 J. W. Scheeren, Recl. Trav. Chim. Pays-Bas, 1986, 105, 71; R. W. Aben
and J. W. Scheeren, Synthesis, 1978, 400; H. Sugimura and K. Osumi,
Tetrahedron Lett., 1989, 30, 1571; S. Vasudevan, C. P. Brock, D. S. Watt
and H. Morita, J. Org. Chem., 1994, 59, 4677; T. Bach, Tetrahedron Lett.,
1994, 35, 5845.
These results suggest that identification of the enantioselec-
tive step can be a complicated problem because the ee will
depend on the relative rates of formation of the various species.
If equilibration between the substrates and oxetanes is much
faster than formation of the Mukaiyama product, the ee will
depend on the relative rates of silyl transfer which lead to the
formation of the two enantiomers of the product. If, however,
the rates of equilibration and product formation are comparable,
the ee will depend on a complex mix of rates associated with all
6 M. Guette, J. Capillon and J.-P. Guette, Tetrahedron, 1973, 29, 3659.
7 A. Wissner, Synth. Comm., 1979, 27.
8 T. Oesterle and G. Simchen, Liebigs Ann. Chem., 1987, 693; T. Sakai, T.
Nakamura, K. Fukuda, E. Amano, M. Utaka and A. Takeda, Bull. Chem.
Soc. Jpn., 1986, 59, 3185.
Received in Corvallis, OR, USA, 7th August 1997; 7/05801B
194
Chem. Commun., 1998