catalyst loading was 30 mol % with respect to substrate, but
this can be lowered to 5 mol % without a significant effect
on the enantioselectivity (vide infra). The reaction was
initiated by addition of 10 equiv of cyclopentadiene (2), and
continued for 3 days at 5 °C, after which the cyclopentadiene
was depleted. The catalytic reaction is ligand accelerated;10
in the absence of a ligand, i.e., using only Cu(NO3)2, no
conversion was observed (entry 1). The catalysts based on
the acridine ligands (L1-L4) gave rise to good conversion
and moderate ee values (entries 2-5). Although the enan-
tiomeric preference was always the same, significant varia-
tions in the ee were observed, compared to aza-chalcone (4).
For example, complexes based on ligands L1 and L2 gave
a lower enantiomeric excess, whereas the copper complex
based on ligand L4 gave rise to a significantly higher
enantioselectivity. In contrast, using the copper complexes
of the bipyridine class ligands (L5-L7), comparable results
are seen for aza-chalcone and 1a with respect to both the
enantiomeric excess, i.e., ee values up to 97%, and enan-
tiomeric preference, albeit with low conversion in the case
of ligands L5 and L6. Surprisingly, by using the catalyst
based on L7 90% conversion was obtained, indicating that
this catalyst is much more reactive, with 97% ee of the (+)
enantiomer. The reason for this remarkable difference in
reactivity of Cu-L7 compared to Cu-L6 is still unclear and
subject to further investigation. On the basis of these results,
it was evident that the copper complex of L7 is the catalyst
of choice, and all further experiments were performed with
this catalytic system.
substrates underwent efficient Diels-Alder reactions with
cyclopentadiene, catalyzed by Cu-L7/salmon testes DNA.
The Diels-Alder product was the sole product; no side
products were detected by NMR and HPLC analysis. In all
cases the endo diastereomer was obtained almost exclusively
and with an excellent enantiomeric excess. The best results
were obtained when the R substituent in the enone was an
aromatic group, with ee values ranging from 94% to 98%.
The electronic nature of the substituent on the aryl ring, i.e.,
electron donating or electron withdrawing, or the position
on the aryl ring did not influence the ee significantly (entry
2-6). In the case of R ) p-ClPh and o-BrPh (entry 4 and
5), the conversion was somewhat decreased, most likely
caused by the low solubility of 1c and 1d. When R was a
2-furanyl group, which is a convenient handle for further
functionalization,11 an excellent ee was obtained (entry 7).
Dienophiles 1f-h, which contain alkyl groups or are unsub-
stituted, also gave rise to high enantiomeric excess, i.e., ee
) 80-88%, although the enantioselectivity is somewhat
lower compared to when R is an aromatic group (entries 7,
8, and 10). No conversion was obtained when cyclopenta-
diene was replaced by the more water soluble diene furan.
The loading of Cu-L7 could be reduced to 5 mol %
without affecting the endo:exo selectivity and the enantio-
selectivity (entries 2, 9, and 11).12
To demonstrate that the present reaction is practical from
a synthetic point of view, we performed the DNA-based
catalytic asymmetric Diels-Alder reaction of 1a with 2 on
a 1.0 mmol (210 mg) scale, with a catalyst loading of 5 mol
%. Product 3a was obtained in 70% isolated yield (85%
conversion), and 96% ee after column chromatography. With
use of the procedure of Evans et al.,9c the imidazole of 3a
was first methylated with methyl triflate, followed by
treatment with methanol. Thus, we obtained the correspond-
ing methyl ester 6 in 41% isolated yield (80% conversion
of the Diels-Alder product; see Scheme 2).13 The optical
The substrate scope of the reaction was investigated by
using R,â-unsaturated 2-acyl imidazoles 1b-h (Table 2). All
Table 2. Substrate Scope of the Diels-Alder Reaction with
R,â-Unsaturated 2-Acyl Imidazoles Catalyzed by CuL7(NO3)2/
st-DNAa
Scheme 2. Synthesis of
3-Phenylbicyclo[2.2.1]hept-5-ene-2-carboxylic Acid Methyl
Ester 6 from 3a
entry
R (substrate)
endo:exo
% eeb
1
2c
3
4d
5e
6
7
8
9c
10
11c
Ph (1a)
Ph (1a)
99:1
99:1
99:1
n.d.
96:4
97:4
94:6
>99:1
>99:1
98:2
99:1
97 (+) (2S,3S)
98 (+) (2S,3S)
p-MeOC6H4 (1b)
p-ClC6H4 (1c)
o-BrC6H4 (1d)
2-furanyl (1e)
cyclohexyl (1f)
Me (1g)
Me (1g)
H (1h)
H (1h)
98
96
94
94
88
86
88
80
83
rotation of the product was [R]20D +126.8 (c 0.401, CHCl3).
From comparison with literature values ([R]20 -130.2,
D
(10) Berrisford, D. J.; Bolm, C.; Sharpless, K. B. Angew. Chem., Int.
Ed. Engl. 1995, 34, 1059-1070.
a All experiments were carried out with st-DNA (1.3 mg mL-1, 2 mM
in base pairs), 0.3 mM [Cu(L7)(NO3)2] (30 mol %), 2 mM 1 in 45 mL of
MOPS buffer (20 mM, pH 6.5) for 3 days at 5 °C, unless noted otherwise.
b For the endo isomer. c 1 mM 1, 0.05 mM Cu(L7)(NO3)2 (5 mol %), 1.3
mg mL-1 st-DNA (2 mM in base pairs). d Conversion ∼50%. e Conversion
∼70%.
(11) Sasaki, S.; Hamada, Y.; Shioiri, T. Tetrahedron Lett. 1997, 38,
3013-3016.
(12) DNA concentration was kept at 1.3 mg/mL, which means that the
Cu-L7/DNA base pair ratio in this case was 1:40.
(13) The same yield and purity of 6 was observed without column
chromatography of Diels-Alder product 3a.
Org. Lett., Vol. 9, No. 18, 2007
3649