Following this strategy, we en-
visioned that cyclic ene-
AHCTUNGTRENGN(UN thio)ureas 6, incorporating an
the backbone 3,3’-substituents on catalyst 5 and the NH
function of 6a are essential for the success of the process.
To reinforce the hydrogen-bonding network and the enan-
tioselectivity of the reaction, enethiourea 6b, having a more
acidic NH proton, was engaged in the Povarov reaction. To
our delight, the reaction of 6b with 3a and 2a in the pres-
ence of 0.1 equivalents of phosphoric acid 5g afforded the
endo-diastereomer 7b as a single detectable stereoisomer in
high yield (81%) and excellent enantioselectivity (92% ee,
entry 2).
NH donor group, might re-
store the missing hydrogen
bond with the catalyst 5,
therefore reestablishing the
high enantioselectivity in the
Povarov reaction via the tran-
sition state shown in the
Figure 2. In addition, we ex-
pected a better diastereoselec-
Figure 2. Hypothesized
dual
activation of an eneurea (ene-
thiourea) and an imine by
a phosphoric acid.
Having established the optimal conditions, we next exam-
ined the scope of this three-component Povarov reaction.
As shown in Table 2, aliphatic aldehydes were appropriate
substrates for this enantioselective three-component synthe-
sis of hexahydropyrroloquinolines 7. Indeed, linear and a-
or b-branched aliphatic aldehydes provided the desired cy-
cloadducts 7 in good yields with high diastereo- and enantio-
selectivities (entries 1–4 and 12–16). It is worth noting that
the competitive isomerization of the in situ formed aliphatic
N-arylimines to the corresponding enamines and/or the par-
tial isomerization of 6b were never observed.[22,23] This cata-
lytic process was also compatible with a,b-unsaturated alde-
hydes (entry 5), electron-neutral, -rich, and -poor aromatic
aldehydes (entries 6–10), and heteroaromatic aldehydes
(entry 11), affording the corresponding products 7g–7m in
similarly good yields (57–83%), high diastereoselectivities
(d.r.>9:1), and excellent enantioselectivities (84–98% ee).
Furthermore, a wide range of diversely functionalized ani-
lines (entries 12–16) were suitable partners for this reaction,
the best enantioselectivity being obtained for electron-poor
substrates (R2 =4-CF3, 96% ee, entry 15). When a meta-sub-
stituted aniline, such as 3-iodoaniline (entry 16), was used,
only the regioisomer 7-iodo-7s was isolated in 89% yield,
without a trace of the sterically more congested 5-iodo
isomer.[14] As expected, the R3 group of the thiourea 6 can
also influence the enantioselectivity of the reaction. While
the reaction with N-3,5-bis(trifluoromethyl)phenyl ene-
thiourea 6c afforded the product with a slightly reduced
enantiomeric excess (entries 2 vs. 1), no reaction occurred
with the bulky N-tert-butyl enethiourea 6d (result not
shown).
The absolute configuration of 7i was unambiguously as-
signed to (2S,3S,4S) by X-ray crystal structure analysis (see
the Supporting Information).[24] Mechanistically, the Povarov
reaction can proceed through either a concerted or a step-
wise process.[25] When using deactivated anilines as reaction
partners, the uncyclized Mannich adduct, in which the inter-
mediate iminium ions are trapped by water, has been isolat-
ed.[26] These results bore evidence of a stepwise mechanism
of the present Povarov reaction. Based on these observa-
tions, a simplistic mechanistic rationale that accounts for the
observed stereoselectivity was proposed (Scheme 2).[27] Ini-
tially, the bifunctional catalyst 5g activates both the imine
and the cyclic enethiourea by hydrogen bonding (compare
Figure 2). A pseudo-intramolecular Si-face attack of the
enethiourea to the iminium ion takes place via the transition
state 8, leading to the N-acyl iminium ion 9,[28] which then
tivity by using 6 instead of 4a, according to the seminal con-
tribution of Batey and co-workers.[12e]
Following this consideration, the three-component reac-
tion of 4-azidobutanal (2a), ethyl 4-aminobenzoate (3a),
and cyclic eneurea 6a in CH2Cl2 at À308C in the presence
of catalytic amounts of a phosphoric acid 5 (0.1 equiv) was
examined (Table 1). After having screened several diversely
Table 1. Survey of phosphoric acid catalysts 5 for the three-component
synthesis of hexahydropyrroloquinolines.[a]
Entry R1
6
Yield [%][b] endo/exo[c] ee [%][d]
1
2
3
4-ClC6H4 [H8] (5a)
6a
6a
6a
6a
6a
6a
48
34
61
41
56
56
72
6
5:1
4:1
5:1
4:1
5:1
27
5
9
42
29
20
88
43
92
4-OMeC6H4 (5b)
4-tBuC6H4 (5c)
SiPh3 (5d)
4
5
6
CH(Ph)2 (5e)
9-anthryl (5 f)
3.5:1
5:1
>95:5
>95:5
7
8
10
2,4,6-(iPr)3C6H2 (5g) 6a
2,4,6-(iPr)3C6H2 (5g) 4a
2,4,6-(iPr)3C6H2 (5g) 6b
81
[a] Reaction conditions: aldehyde 2a (0.1 mmol), aniline 3a (0.1 mmol),
ene(thio)urea (0.15 mmol), and catalyst (0.01 mmol) in CH2Cl2
A
6
5
(1.0 mL) at À308C. [b] Isolated yield after column chromatography.
[c] Ratio of endo/exo stereoisomers was determined by 1H NMR spec-
troscopy. [d] Enantiomeric excess was determined by HPLC analysis on
a chiral stationary phase.
substituted chiral phosphoric acids 5, we found that the
bulky (R)-TRIP catalyst (5g, TRIP=3,3’-bis(2,4,6-triisopro-
pylphenyl)-1,1’-binaphthyl-2,2’-diyl hydrogenphosphate) was
optimal in terms of both yield (72%) and enantiomeric
excess of the product 7a (88% ee), although the diastereose-
lectivity remained unsatisfactory (entry 7).[21] In a control
experiment using cyclic enecarbamate N-Cbz-2,3-dihydro-
pyrrole (4a) as dienophile, the corresponding cycloadduct
was obtained in a much lower yield (6%) and lower enan-
tiomeric excess (43% ee) under otherwise identical reaction
conditions (entry 8). These results clearly indicate that both
5870
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5869 – 5873