5
mechanistic studies. Last but not least, many catalysts are
only accessible by complicated multistep reaction sequences
6
or need rare and expensive metals as reactive centers. Recent
research is focused on the development of stable and storable
catalysts, and the first remarkable results have been presented
7
by Kobayashi et al., who presented stable rare-earth and
zirconium catalysts for Mannich, aza-Diels-Alder, and
Mukaiyama aldol reactions, and Bull et al., who described a
moisture-stable titanium triflate for aza-Diels-Alder reac-
tions.8
During our ongoing studies of the mechanisms of titanium-
(IV)-alkoxide-mediated direct aldol additions, we explored
the catalytic potental of Mikami’s tetranuclear titanium
9
cluster. Mikami’s catalyst is crystalline and stable even
against boiling 1 N HCl and 1 N LiOH in dioxane and easy
i
to synthesize from Ti(O Pr)
4
, R-BINOL, or S-BINOL and
6
10
water. We also synthesized this cluster with rac-BINOL
and found that all six BINOL molecules which are incor-
porated in each cluster have the same stereochemistry. No
clusters appeared in crystalline form with a mixed stereo-
chemistry of the incorporated BINOLs. Therefore, the
clusters obtained from rac-BINOL are identical to those
produced with R-BINOL or S-BINOL (Figure 1).
Figure 1. Crystal structure of the tetranuclear titanium complex
rac-Ti (µ-BINOLato) (µ -OH) ; hydrogen atoms are omitted for
clarity.
4
6
3
4
13
ketone. In many cases, only one regioisomer was obtained.
Methyl groups of alkan-2-ones were found to be unaffected
under these reaction conditions. Therefore, the Wieland-
Miescher ketone 4 avoids the typically aldol cyclization to
the CD bicyclic steroidal intermediate 6 (Scheme 1).14 The
To test the applicability of this titanium complex, we
reacted several aldehydes with symmetrical and unsym-
metrical ketones. In fact, the catalyst was able to promote
the direct aldol addition between aldehydes 1 and ketones 2
in a remarkably clean way even with very low catalyst
11
12
loadings to give the aldols 3 (Table 1). This reaction was
found to be highly regioselective. Aldol addition is strongly
preferred at the sterically more encumbered R-side of the
Scheme 1. Regioselective Formation of Diketone 5
(
5) (a) Keck, G. E.; Krishnamurthy, D. J. Am. Chem. Soc. 1995, 117,
2
3
363-2364. (b) Kitamoto, D.; Imma, H.; Nakai, T. Tetrahedron Lett. 1995,
6, 1861-1864. (c) Terada, M.; Matsumoto, Y.; Nakamura, Y.; Mikami,
K. Inorg. Chim. Acta 1999, 267-272. (d) Delas, C.; Szymoniak, J.; Lefranc,
H.; Mo ¨ı se, C. Tetrahedron Lett. 1999, 40, 1121-1122. (e) De Rosa, M.;
Soriente, A.; Scettri, A. Tetrahedron: Asymmetry 2000, 11, 3187-3195.
(
f) Davis, J. T.; Balsells, J.; Caroll, P. J.; Walsh, P. J. Org. Lett. 2001, 3,
6
99-702. (g) Zimmer, R.; Peritz, A.; Czerwonka, R.; Schefzig, L.; Reissig,
H.-U. Eur. J. Org. Chem. 2002, 3419-3428. (h) Villano, R.; De Rosa, M.;
Salerno, C.; Soriente, A.; Scettri, A. Tetrahedron: Asymmetry 2002, 13,
1
949-1952. (i) Mahrwald, R. Org. Lett. 2000, 2, 4011-4012. (j) Villano,
R.; Acocella, M.; De Rosa, M.; Soriente, A.; Scettri, A. Tetrahedron:
Asymmetry, 2004, 15, 2422-2424;
(6) (a) Sawamura, M.; Ito, Y. In Catalytic Asymmetric Synthesis, 2nd
ed.; Ojima, I., Ed.; Wiley-VCH: New York, 2000. (b) Ito, Y.; Sawamura,
M.; Hayashi, T. J. Am. Chem. Soc. 1986, 108, 6405-6406. Ito, Y.;
Sawamura, M.; Hayashi, T. Terahedron Lett. 1987, 28, 6215-6218. (c)
Sawamura, M.; Ito, Y. Catal. Asymmetric Synth. 1993, 367-388.
bicyclic dione 5 is formed and strongly preferred with a high
degree of diastereoselectivity (>95:5). Dione 5 posseses two
(
7) (a) Ishikawa, S.; Hamada, T.; Manabe, K.; Kobayashi, S. J. Am.
Chem. Soc. 2004, 126, 12236-12237. (b) Kobayashi, S.; Ueno, M.; Mizuki,
Y.; Ishitani, H.; Yamashita, Y. Proc. Natl. Acad. Sci. U.S.A. 2004, 101,
5
476-5481.
8) Bull, S. D.; Davidson, M. G.; Johnson, A. L.; Robinson, D. E. J. E.
Chem. Commun. 2003, 1750-1751.
9) Mikami, K.; Ueki, M.; Matsumoto, Y.; Terada, M. Chirality 2001,
3, 541-545.
10) The experimental procedure was the same as that given by Mikami
(
(11) So far, the catalyst loadings for similar reactions range usually
between 2 and 20 mol %; recently, Ding et al. presented an enantioselective
carbonyl-ene reaction with comparable low catalyst loadings: Yuan, Y.;
Zhang, X.; Ding, K. Angew. Chem. 2003, 42, 5478-5480.
(
1
(
(12) Typical experimental procedure: 1 equiv of aldehyde and 1.5 equiv
of ketone were mixed at room temperature. If problems with the solubility
occurred, small amounts of CH2Cl2 were used as solvent. A portion 0.2
mol % of the catalyst was added. The procedure of the reaction was
monitored by TLC, and when the turnover was complete, the reaction
mixture was diluted with diethyl ether; the reaction was then quenched with
aqueous NH4Cl. The organic layer was separated, dried (MgSO4), and
filtered, and the ether was removed in vacuo. CC (hexane/ethyl acetate)
afforded the pure aldols. Yields have not been optimized. Large amounts
of recovered aldehyde indicate the necessity of the employment of longer
reaction times or higher reaction temperatures in some cases.
et al. The catalyst was obtained as dark red rectangular crystals. A single-
crystal structure analysis was made of a suitable crystal. The complex was
found to crystallize in the cubic space group F-43c. a ) b ) c ) 32.348
Å; R1 ) 0.049, wR2 ) 0.118. Crystallographic data have been deposited
at the Cambridge Crystal Data Center (CCDC 279234). This material can
be obtained upon request to CCDC, 12 Union Road, Cambridge, 1EZ, U.K.
(http://www.ccdc.cam.ac.uk; e-mail at deposit@ccdc.cam.ac.uk). The struc-
ture was refined with SHELX97 (Sheldrick, G. M. SHELX97: Program
for crystal structure refinement; Universit a¨ t G o¨ ttingen: G o¨ ttingen, Ger-
many).
282
Org. Lett., Vol. 8, No. 2, 2006