atom also leads to decreased selectivity. The chemical
efficiencies using the carbamoyl ligands (3-8) were good
(>80%) with little variation from ligand to ligand. The
dependency of ee on variation in the alkyl group is not clear.
The ligand 9 with the free NH group decreased the reactivity
of the catalyst significantly. Modifying the type of substituent
on the nitrogen can modulate the steric bulk and electronic
nature of the ligand. Changing the carbamoyl group on
nitrogen to a urea (10), sulfonamide (11), or amide (12)
resulted in decreased reactivity and selectivity.16 Large groups
on the tertiary alcohol gave high selectivity (3 and 14),
whereas the small methyl group (13) decreased selectivity.
Lanthanides, due to their larger size and vacant f-orbitals,
can attain higher coordination numbers than the Lewis acids
in either main group or transition metals. Such higher
coordination numbers can be achieved either by varying the
stoichiometry of the Lewis acid to ligand or by introducing
achiral additives (Scheme 2, Table 2). Increasing the amount
Table 2. Effect of Additives
entry
additive (equiv)
yield (%)a
ee (%)b
1
2
3
none
84
72
86
84
73
84
67
63
95
67
45
63
79
64
58
49d
64
82
89
89
84
73
83
92
3 (1 equiv)c
3 (2 equiv)c
4
5
6
(()-2 (2 equiv)
HOCH2CH2OH (2 equiv)
15 (1 equiv)
7
15 (2 equiv)
8
15 (3 equiv)
9
16 (2 equiv)
10
11
12
MS 4 Å (17 mg)
MS 4 Å (150 mg)
15 (2 equiv) + MS 4 Å (150 mg)
a Isolated yields. The amount of Lewis acid used was 30 mol %. b HPLC
analysis was used to determine ees. c Amount in parentheses refers to the
extra ligand added. d See Supporting Information for calculation of the ee.
an additive since it had a beneficial effect in diastereo-
selective radical reactions mediated by lanthanide Lewis
acids.19 In contrast, ethylene glycol as an additive in
enantioselective transformations was not favorable (compare
entry 1 with 5). Next, we evaluated two N-acyl oxazolidi-
nones (15-16) as additives.20 Of these, N-benzoyl oxazoli-
dinone 16 was the most effective (compare entry 1 with
entries 7 and 9). Also, a dependence of selectivity on the
amount of additive was observed: 2 equiv of the additive
with respect to the chiral Lewis acid was found to be optimal
(compare entries 6-8). We believe that after the 2 equiv is
added, there is no vacant coordination site in the reactive
complex for the additive to bind to the Lewis acid.
Scheme 2
of the ligand to 2 or 3 equiv compared to the Lewis acid
decreased the enantioselectivity (compare entry 1 with entries
2 and 3). In this reaction, the product due to its similarity
with the substrate can coordinate to the reactive complex.17
To test this possible influence of the product, 2 equiv
(compared to the Lewis acid) of the (()-product 2 was added,
and the ee of the product derived from the reaction decreased
to 49% (entry 4). To prevent the product from hampering
the efficacy of the chiral catalyst, achiral additives were
considered. Achiral additives are known to affect the
selectivity in asymmetric catalysis.18 This is due to their
coordination to the chiral Lewis acid and hence alteration
of the catalyst superstructure and also to their filling of empty
coordination sites. We initially evaluated ethylene glycol as
Molecular sieves play a major role in catalysis. Although
the exact nature of their effect is not well understood, it is
widely believed that they either remove the adventitious
water in the catalyst or aid in blocking coordination sites.21
Also, the amount of molecular sieves is an important
parameter for observing beneficial effects.10d In our case,
addition of small amounts of MS 4 Å led to a decrease in
enantioselectivity (compare entry 1 with entry 10). However,
larger amounts (150 mg) of MS 4 Å produced a small
enhancement in selectivity (compare entry 1 with entry 11).
Finally, we were able to obtain 92% enantioselectivity by
combining 2 equiv of the additive 16 and MS 4 Å with 30
mol % of the chiral Lewis acid (entry 12).
(12) 68% ee (10 mol %); 74% ee (20 mol %); 79% ee (30 mol %); 64%
ee (40 mol %).
Assuming that two triflate ions are bound to the metal
and that the substrate, ligand, and additive bind in a bidentate
(13) (a) Schaus, S. E.; Jacobsen, E. N. Org. Lett. 2000, 2, 1001-1004.
(b) Kobayashi, S.; Hamada, T.; Nagayama, S.; Manabe, K. Org. Lett. 2001,
3, 165-167. (c) Inanaga, J.; Sugimoto, Y.; Hanamoto, T. New J. Chem.
1995, 19, 707-712. (d) Evans, D. A.; Nelson, S. G.; Gagne, M. R.; Muci,
A. R. J. Am. Chem. Soc. 1993, 115, 9800-9801.
(14) See Supporting Information in ref 6a for details on stereochemical
analysis.
(19) Examples of ether/alcohol/carboxylic acid/amine-type additives in
catalysis with lanthanides: (a) Sibi, M. P.; Rheault, T. R. J. Am. Chem.
Soc. 2000, 122, 8873-8879. (b) Aspinall, H. C.; Bissett, J. S.; Greeves,
N.; Levin, D. Tetrahedron Lett. 2002, 43, 319-321. (c) Lacoˆte, E.; Renaud,
P. Angew. Chem., Int. Ed. 1998, 37, 2259-2262. (d) Fukuzaka, S.-I.; Seki,
K.; Tatsuzawa, M.; Mutoh, K J. Am. Chem. Soc. 1997, 119, 1482-1483.
(20) Shibasaki, M.; Yamada, K.-I.; Yoshikawa, N. In Lewis Acids in
Organic Synthesis; Yamamoto, H., Ed.; Wiley-VCH: Weinheim, 2000; pp
911-944 and references therein.
(15) Bhaskar Kanth, J. V.; Periasamy, M. Tetrahedron 1993, 49, 5127-
5132.
(16) Reactions with these ligands were considerably slower. Yields
averaged around 70%.
(17) For a good evaluation/analysis of product influence in asymmetric
Diels-Alder reaction, see: Heller, D. P.; Goldberg, D. R.; Wulff, W. D. J.
Am. Chem. Soc. 1997, 119, 10551-10552.
(18) Review: Vogl, E. M.; Groger, H.; Shibasaki, M. Angew. Chem.,
Int. Ed. 1999, 38, 1570-1577.
(21) (a) Aspinall H. C.; Dwyer, J. L.; Greeves, N.; McIver, E. G.; Wooley,
J. C. Organometallics 1998, 17, 1884-1888. (b) Kodama, H.; Ito, J.; Hori,
T.; Furukawa, I. J. Organomet. Chem. 2000, 603, 6-12. (c) ref 10d.
Org. Lett., Vol. 4, No. 17, 2002
2931