Acid-promoted aza-cyclization versus -cyclization of N -acyliminium species into fused pyrrolo[1,2-a]imidazolones and pyrrolo[2,1- a ]isoquinolinones

Article abstract of DOI:10.1055/s-0030-1260949

A new approach for the synthesis of fused imidazolones and isoquinolinones is presented. The key step of this sequence was the interception of an N-acyliminium species with nitrogen or -aromatic nucleophiles under kinetic vs. thermodynamic control. In add

Full text of DOI:10.1055/s-0030-1260949

LETTER
1821
Acid-Promoted Aza-Cyclization versus p-Cyclization of N-Acyliminium
Species into Fused Pyrrolo[1,2-a]imidazolones and Pyrrolo[2,1-a]-
isoquinolinones
A
cid-Prom
e
oted
A
za-
C
a
yclization
n
versus
p
-Cyc
-
lization
F
of
N
-Acylimi
r
nium
S
p
a
e
cies nçois Fleury, Pierre Netchitaïlo, Adam Daïch*
Laboratoire de Chimie, URCOM, EA 3221, INC3M CNRS-FR 3038, UFR des Sciences et Techniques, Université du Havre,
BP: 540, 25 Rue Philippe Lebon, 76058 Le Havre Cedex, France
Fax +33(2)32744391; E-mail: adam.daich@univ-lehavre.fr
Received 7 February 2011
In both cases, Brønsted and Lewis acids were used as pro-
moters.
Abstract: A new approach for the synthesis of fused imidazolones
and isoquinolinones is presented. The key step of this sequence was
the interception of an N-acyliminium species with nitrogen or p-ar-
omatic nucleophiles under kinetic vs. thermodynamic control. In
addition, in the presence of two p-aromatic nucleophiles, only the
six-membered ring closure into pyrroloisoquinolinones occurred.
Moreover, cationic cyclization using a nitrogen atom as
internal nucleophile was first mentioned during the syn-
thesis of 9a-phenyltetrahydroisoindolo[2,1-a]quinazo-
line-5,11-dione6a and ten years later for the production of
diethyl tetrahydrodioxolo[4,5-g]pyrrolo[2,1-b]quinazo-
line-8,8-dicarboxylate.6b Its power was also illustrated in
the total synthesis of ( )- and (–)-physostigmine,7 ( )-
glochidine, and ( )-glochidicine,8 imidazoloindoline al-
kaloids,9a oroidin-derived alkaloids,9b and more recently,
the bioactive enantiopure (–)-decarbamoyloxy-saxitox-
in.10 Reports on this process concern also its use to access
aldose reductase inhibitors,11 in parallel synthesis of Δ2-2-
oxopiperazines,12 in asymmetric synthesis of Reissert
compounds,13 in synthesis of lactam-based peptidomimet-
ics,14 and in sterically hindered N-substituted and fused
g-lactams.15
Key words: N-acyliminium, diastereoselectivity, aza-cyclization,
pyrrolo[1,2-a]imidazolone, pyrrolo[2,1-a]isoquinolinone
N-Acyliminium species are intermediates which are wide-
ly used in modern organic synthesis. Their importance is
underlined by numerous reviews of their generation and
use for the elaboration of natural and unnatural aza-het-
erocyclic systems with therapeutic interest.1 If the forma-
tion of the C–C bonds by this protocol in inter- and
intramolecular fashion is widely applied by the scientific
community, the one which consists in the formation of the
C–X linkage with X as a heteroatom (X = O, S, Se, and N)
appears to be a new field of application of this chemis-
try.2,3
To the best of our knowledge, the use of the present appli-
cation in both intramolecular and diastereoselective N-
acyliminium-mediated cyclization represents a novel il-
lustration of this chemistry (Scheme 2). This can result in
the formation of a central five-membered ring of N,N-ac-
etal 3 by nitrogen atom attack of the cation II (path a).
However, when the aromatic system at the nitrogen a-po-
sition of II is sufficiently activated, a six-membered ring
of fused isoquinoline 4 (path b) can be obtained. The cores
of these tricyclic systems ars often encountered in natural
products and biologically active compounds.16
We have described earlier the potential for these species
to intercept intramolecularly nitrogen nucleophiles in
acidic media.2 In particular, their presence with both car-
bon- and nitrogen-tethered nucleophiles showed the reac-
tion operability under kinetic vs. thermodynamic control
(Scheme 1). This resulted in the formation of C–N bond
(1, reversible) or C–C bond (2, irreversible) depending on
the reaction conditions.3 Similar processes in oxygen se-
ries are relatively well documented to provide, in particu-
lar, asymmetric macrocyclic N,O-acetals4 and alkaloids.5
For this purpose, the requisite a-hydroxy lactam 9 was ob-
tained in three steps from commercially available (R)-
phenylglycinol (5, Scheme 3). Protection of (R)-5 with
O
O
O
R1
R1
R1
R2
fast
a
slow
b
N
N
N
N
R2
R2
a
NH
HN
Ph
N-acyliminium I
b
Ph
1
2
Scheme 1 Representative kinetic vs. thermodynamic control of cationic cyclization of N-acyliminiums of type I
SYNLETT 2011, No. 13, pp 1821–1826
x
x
.x
x
.2
0
1
1
Advanced online publication: 14.07.2011
DOI: 10.1055/s-0030-1260949; Art ID: D04011ST
© Georg Thieme Verlag Stuttgart · New York

1822
J.-F. Fleury et al.
LETTER
O
O
O
!
!
fast
slow
N
N
N
b
a
NRPg
a
Ph
N
RN
Pg
Ph
Pg
b
3
N-acyliminium II (R = H, amide)
4
Scheme 2 Plausible kinetic vs. thermodynamic control of cationic cyclization of N-acyliminiums of type II
(Boc)2O provided N-Boc derivative (R)-6 in 73% yield ac- by using selective NOE difference measurements. In the
cording to the known procedure.17 The second step of our major product 3b a strong NOE effect was observed con-
sequence consisted in coupling the imide 7 with N-Boc- firming the cis orientation of H2 and H10 (Scheme 4).21
(R)-phenylglycinol (6). The Mitsunobu reaction was im- This showed also that the stereochemical outcome during
plemented with DIAD instead of DEAD usually em- the formation of the major isomer is similar to that we re-
ployed as coupling reagent by applying conditions of ported earlier in the oxygen series.22
Martinez et al. (imide 7, DIAD, Ph3P, THF, 0 °C).18 Un-
der these conditions, the expected chiral imide 8 was iso-
lated in 79% yield.
Taking into account that TFA is a standard deprotecting
agent for the Boc group, we decided to use it for the cy-
clization process, hoping to be able to promote both the
It has been demonstrated1 that the reduction of one of both cyclization reaction and the amine deprotection at the
imide carbonyls can lead to several alcohols including the same time. Thus, a-hydroxy lactam 9 (each isomer or both
secondary a-hydroxy lactam we required as N-acylimini- diastereomers) upon treatment with neat TFA for 30 min-
um precursor. After optimization work, the chemoselec- utes at room temperature followed by alkaline hydrolysis
tive reduction of (R)-8 was performed with a slight excess gave isoquinolines 4a and 4b. These tricyclic systems
of LiEt3BH (1.15 equiv) in analogy with our reports.19 were isolated as free amines in an inseparable 2:1 ratio
This afforded the a-hydroxy lactam 9 (78% yield) as a and excellent 95% yield. Finally, by using an excess of
mixture of two diastereomers (7:1) separable by chroma- BF3⋅OEt2 according to the standard protocol (2.5 equiv
tography (EtOAc–cyclohexane = 2:3).
BF3⋅OEt2, CH2Cl2, 0 °C), after one hour of the reaction at
room temperature we also observed the formation of the
unique isoquinoline system present as only one diastereo-
mer 4a accompanied by the deprotection of the amine
function. Interestingly, the latter fact has rarely been ob-
served under BF3⋅OEt2 influence and consequently little
reported in the literature.
According to previous reports1 demonstrating that
Brønsted and Lewis acids are good catalysts for various
amidoalkylations, we thought to use PTSA, TFA, and
BF3·OEt2 for the intramolecular p-cyclization and/or a-
hetero-amidoalkylation in order to measure the impact of
these acids on the ring-closure step. Thus, treatment of 9
(two diastereomers), as a precursor of stable cation II The formation of both isomers 4a,b can be explained by
(Scheme 2; Scheme 4), with PTSA (10 mol%) in dry the p-aromatic attack on either face of the cation II
CH2Cl2 at reflux for 30 minutes afforded N-Boc-imidazo- formed in acid medium by invoking a Felkin–Ahn-type23
isoindolone diastereomers 3a and 3b in 5.1:4.9 ratio.20 model (Scheme 4). The plausible chelation in the cation II
This mixture, which resulted from exclusive a-aza-ami- is probably at the origin of the diastereoselectivity we ob-
doalkylation, was obtained in quantitative yield and sepa- served (4a as major isomer or exclusive isomer). Further-
rated simply by silica gel chromatography. The more, the formation of the imidazoisoindolones 3 and
stereochemical relationship in 3a and 3b was established isoquinolines 4 seems to occur by kinetic versus thermo-
OH
OH
i) THF, 0 °C
imide 7 (0.8 equiv)
Boc2O (1.1 equiv)
MeCN, r.t.
O
NH2
NHBoc
ii) DIAD (1.1 equiv)
iii) Ph3P (1.5 equiv)
iv) r.t., 4 h
3 h
N
73%
O
NHBoc
(R)-6
(R)-5
(R)-8, 79%
HO
N
O
i) LiEt3BH (1.15 equiv)
CH2Cl2, –78°C, Ar, 15 min
ii) H2O
N H
O
O
78%
7
NHBoc
2 diastereomers (7:1)
separable
9
Scheme 3 Sequence leading to the N-acyliminium precursor 9
Synlett 2011, No. 13, 1821–1826 © Thieme Stuttgart · New York

LETTER
Acid-Promoted Aza-Cyclization versus p-Cyclization of N-Acyliminium Species
1823
2 diastereomers
(5.1:4.9), separable
2 diastereomers
(2:1), inseparable
O
O
trans
N
N
N
NH2
HO
Boc
N
ii
95%
O
3a
i
4a,b
NHBoc
+
>99%
O
O
iii
5
89%
N
cis
N
9
10
NH2
2
H
N
2 diastereomers
(7:1), separable
trans
H
Boc
NOE
7–9%
formed as a sole
diastereomer
4a
3b
H
stereogenic
center
H
O
H
N
N
N
H
O
HN
H
O
N
NHBoc
Boc
Boc
chelation of N-acyliminium
N-acyliminium
intermediate II
Felkin–Ahn model
type
intermediate II
Scheme 4 Cyclization reaction. Reagents and conditions: (i) PTSA (10 mol%), CH2Cl2, reflux, 30 min; (ii) (a) TFA, r.t., 30 min; (b) aq
NaHCO3, CH2Cl2; (iii) (a) BF3·OEt2 (2.5 equiv), CH2Cl2, 0 °C; (b) r.t., 1 h.
dynamic control using the formal species II as intermedi- In another approach, the competing electrophilicity of the
ate. This was confirmed when 3a,b as a mixture of two N-acyliminium species towards p-aromatic systems was
diastereomers was treated with neat TFA under conditions again explored. Herein, the principal objective was to ac-
ii as in Scheme 4 but for a longer time (4 h). Under these cess the isoindolo[1,2-a]isoquinolin-8(12bH)-one skele-
conditions, imidazoisoindolones 3a,b generate the cation- ton starting from known keto acid 10 (Scheme 5) but
ic platform II, and the isoquinolines 4a,b were again ob- substituted at the amine group at C5, hoping to be able to
tained in 90% yield and comparable ratio of 2.5:1.5 with separate the diastereomers.
the same stereodistribution as before.
As outlined in Scheme 5, the first key substrate (S,R)-12
was reached in two steps. Thus, the known enantiopure
oxazolidine product (R,R)-11, prepared by Meyers’ meth-
od in 75% yield,24 in the presence of TiCl4 and Et3SiH in
dry CH2Cl2 provided the expected amido alcohol (S,R)-12
in 85% yield and high diastereoselectivity since only one
diastereomer was isolated. The usual rationalization25 in-
vokes initial formation of the intermediate I. Further, the
attack of the hydride occurs at the cationic position on the
upper side of the Felkin–Ahn model such as in the inter-
mediate K which would be favorably formed by a chela-
tion process in preference to cation J.
O
O
CO2H
O
R
R
N
R
N
S
i
ii
75%
ref. 24
85%
O
OH
10
11
12
OTiCl3
N
O
H
H
Amido alcohol 12 was then engaged in the Mitsonubu re-
action according to the protocol outlined above for the
transformation of (R)-6 into 8 (Scheme 3). Under these
conditions, the enantiopure imide 13 was isolated in 52%
yield after purification by chromatography (Scheme 6).
N
O
H
N
O
OTiCl3
OTiCl3
J (unfavored)
K (favored)
I
The imide 13 was then reduced chemoselectively as above
and led, after basic hydrolysis, to pure hydroxy lactam 14
in 87% yield. This product, obtained as a mixture of two
separable diastereomers 14a,b (3:2 ratio), results from the
Scheme 5 Synthesis of alcohol 12. Reagents and conditions: (i) (R)-
5 (1 equiv), toluene, DS, reflux, 12 h; (ii) (1) Et3SiH (3.2 equiv),
CH2Cl2, –78 °C; (2) TiCl4 (2 equiv), –78 °C, 1 h 30; (3) r.t., 12 h; (4)
sat. aq NH4Cl, r.t., 30 min.
Synlett 2011, No. 13, 1821–1826 © Thieme Stuttgart · New York

1824
J.-F. Fleury et al.
LETTER
O
i
ii
O
N
S
OH
R
R
R
52%
87%
*R
*R
N
N
OH
O
O
O
13
12
14a,b (3:2)
separable
N
S
R* =
iii
Nu2
O
O
Nu1
N
N
R*
R*
+
66%
*R
N
Nu1
Nu1
cis
trans
15a,b (1:1 then 2.5:1) separable
O
L
R*
H
N
H
H
H
H
TS1
TS2
H
15a (cis)
major
R*
15b (trans)
minor
N
O
O
TS1 more stable than TS2
O
Nu1
Nu2
O
9
8
I
N
6
N
N
I
O
II
12b
1
N
O
N
N
O
12
5
4
Nu1
Nu2
O
II
15a,b
16a,b
intermediate L
Scheme 6 Reagents and conditions: (i) (1) phthalimide (7, 1.1 equiv), THF, (2) DIAD (1.2 equiv), 0 °C; (3) Ph3P (1.5 equiv), 0 °C; (4) r.t.,
2 h; (ii) (1) LiEt3BH (1.15 equiv), CH2Cl2, –78 °C, 15 min; (2) sat. aq NaHCO3, r.t.; (iii) (1) TFA, Ar, r.t., 2 h; (2) CH2Cl2, NaHCO3.
attack of the hydride on both faces of the phthalimide nu- when the reaction was refluxed for two hours in the same
cleus. In addition, the observed relative selectivity in the solvent with other parameters unchanged. The results
reaction can be explained by the presence of a stereogenic show that the stereochemical distribution of products
center at the b-position of the phthalimide nucleus which 15a,b,27 separable by chromatography, was now of 2.5:1
directs the approach of the hydride.
ratio in favor of the diastereomer 15a which could be con-
sidered as a thermodynamically favored isomer. Never-
theless, in spite of this success, the reaction yield (48% of
the crude reaction products) was lower than the one ob-
tained above (66%), and the reaction was associated with
decomposition of the reactants in acid medium. Since the
configuration of both compounds could not be determined
by X-ray crystallography analysis due to the difficulties
encountered during recrystallization attempts, it was ten-
tatively assigned on the basis of studies given by the
Speckamp group and initiated by the Maryanoff et al. on
pyrrolo[2,1-a]isoquinolinones substituted with alkyl or
aryl groups at their C6-position.28 In these cases, they ob-
served an isomeric ratio for 6a/6b ranging from 55.5:34.5
up to 93:7 depending on the nature of the substituent. The
stereochemical outcome of the reaction has been inter-
preted in terms of the steric requirements of the substitu-
ent at C6 favoring a chairlike transition state as the
cyclization intermediate. In this case, the more stable con-
Each cation precursor 14a or 14b, having been separated
chromatographically, was alternatively subjected to the
intramolecular p-cyclization under conditions outlined
above (Scheme 6). This resulted in the formation of the
presumably stable cation L which undergoes an intramo-
lecular p-amidoarylation. Interestingly, after two hours of
the reaction, in both cases approximately 66% of diaste-
reomeric mixture of products 15a,b with very poor diaste-
reoselectivity (1:1 ratio) was isolated. This shows that
exclusively the expected six-membered-ring pathway is
effective according to the Baldwin rules (path I,
Scheme 6).26 Similarly, starting from crude hydroxy lac-
tam 14a,b (3:2 ratio) without purification and separation,
the reaction likewise led to the same mixture in compara-
ble yield and stereochemical distribution.
Importantly, it was found that the p-cyclization of 14a
and/or 14b into 15a,b could be improved in some aspects,
Synlett 2011, No. 13, 1821–1826 © Thieme Stuttgart · New York

LETTER
Acid-Promoted Aza-Cyclization versus p-Cyclization of N-Acyliminium Species
1825
1992, 57, 1568. (b) Yamagishi, M.; Yamada, Y.; Ozaki, K.;
Asao, M.; Shimizu, R.; Suzuki, M.; Matsumoto, M.;
Matsuoka, Y.; Matsumoto, K. J. Med. Chem. 1992, 35, 2085.
(12) Lee, S.-C.; Park, S. B. J. Comb. Chem. 2007, 9, 828.
(13) (a) Surygina, O.; Ehwald, M.; Liebscher, J. Tetrahedron
Lett. 2000, 41, 5479. (b) Sieck, O.; Schaller, S.; Grimme, S.;
Liebscher, J. Synlett 2003, 337.
(14) (a) Le Quement, S. T.; Nielsen, T. E.; Meldal, M. J. Comb.
Chem. 2007, 9, 1060. (b) Min, B. J.; Gu, X.; Yamamoto, T.;
Petrov, R. R.; Qu, H.; Lee, Y. S.; Hruby, V. J. Tetrahedron
Lett. 2008, 49, 2316; and references cited therein.
(15) (a) Sannigrahi, M.; Pinto, P.; Chan, T. M.; Shih, N.-Y.;
Njoroge, F. G. Tetrahedron Lett. 2006, 47, 4877.
formation in TS1 compared with TS2 would lead to cis-
15a as the major product (Scheme 6).29
In conclusion, we have developed a short and effective
synthesis of enantiopure fused pyrroloimidazolones 3 and
pyrroloisoquinolinones 4 based on kinetic vs. thermody-
namic control of an intramolecular electrophilic cycliza-
tion when both nitrogen and p-aromatic nucleophiles are
present in an N-acyliminium precursor. The reaction pro-
ceeds with good to very high yields and moderate diaste-
reoselectivity. Moreover, with two p-aromatic nucleo-
philes, the N-acyliminium cyclization onto chiral substi-
tuted isoindoloisoquinolinones 15 related to 4 occurs with
acceptable yields and low stereochemical distribution.
Only a six-membered-ring pathway as opposed to an
eight-membered occurred during these transformations.
In this case the diastereomeric ratio could be interestingly
modified (up to 2.5:1 ª 71.5:28.5 in non-optimized con-
ditions) by varying the reaction temperature without
changing other parameters.
(b) Cayley, A. N.; Gallagher, K. A.; Ménard-Moyon, C.;
Schmidt, J. P.; Diorazio, L. J.; Taylor, R. J. K. Synthesis
2008, 3846.
(16) Moreau, A.; Couture, A.; Deniau, E.; Grandclaudon, P. Eur.
J. Org. Chem. 2005, 3437; and references cited therein.
(17) Lewanowicz, A.; Lipinski, J.; Siedlecka, R.; Skarzewski, J.;
Baert, F. Tetrahedron 1998, 54, 6571.
(18) Martinez, E. J.; Corey, E. J. Org. Lett. 2000, 2, 993.
(19) Pin, F.; Comesse, S.; Garrigues, B.; Marchalín, Š.; Daïch, A.
J. Org. Chem. 2007, 72, 1181.
(20) The ratio of both diastereomers 3a,b considered as kinetic
products was estimated by 1H NMR spectroscopy and is
different from that of their amidal congeners 9 in a 7:1 ratio.
(21) Data for Compound 3b
These simple modular approaches are currently under in-
vestigation in our group to access various scaffolds anal-
ogous to natural products.
Isolated in 51% yield as a white solid (EtOAc–cyclohexane
= 1:4); mp 136 °C; [a]D –20.9 (c 0.81×10–3, CH2Cl2).
IR (KBr): nmax = 3019, 1687 cm–1. 1H NMR (300 MHz,
CDCl3): d = 1.01 (s, 9 H), 3.33 (dd, 1 H, J = 12.0, 8.0 Hz),
4.64 (dd, 1 H, J = 12.3, 7.6 Hz), 4.80 (t, 1 H, J = 7.8 Hz),
7.25–7.38 (m, 5 HAr), 7.51 (t, 1 HAr, J = 7.2 Hz), 7.58 (t,
1 HAr, J = 7.2 Hz), 7.80 (d, 1 HAr, J = 7.2 Hz), 8.16 (d, 1 HAr,
J = 7.2 Hz). 13C NMR (75 MHz, CDCl3): d = 28.1, 51.7,
64.7, 75.9, 81.0, 122.8, 125.4, 126.8, 127.7, 128.9, 131.8,
132.7, 144.3, 153.6, 179.9. MS (EI): m/z = 350 [M+]. Anal.
Calcd (%) for C21H22N2O3 (350.16): C, 71.98; H, 6.33; N,
7.99. Found: C, 71.77; H, 6.18; N, 7.76.
Acknowledgment
We are grateful to Le Havre University for material support (J.F.F.)
and to Dawn Hallidy for proofreading the text.
References and Notes
(1) For important reviews in this area, see: (a) Speckamp, W.
N.; Moolenaar, M. J. Tetrahedron 2000, 56, 3817.
(b) Maryanoff, B. E.; Zhang, H. C.; Cohen, J. H.; Turchi,
I. J.; Maryanoff, C. A. Chem. Rev. 2004, 104, 1431.
(2) For aza-cyclization, see: (a) Fogain-Ninkam, A.; Daïch, A.;
Netchitaïlo, P.; Decroix, B. Eur. J. Org. Chem. 2003, 427.
(b) Oukli, N.; Comesse, S.; Chafi, N.; Oulyadi, H.; Daïch, A.
Tetrahedron Lett. 2009, 50, 1459.
(22) Sikoraiova, J.; Daïch, A.; Marchalín, Š.; Decroix, B.
Tetrahedron Lett. 2002, 43, 4747.
(23) (a) Polniaszek, R. P.; Belmont, S. E. J. Org. Chem. 1991, 56,
4868. (b) Chihab-Eddine, A.; Daïch, A.; Jilale, A.; Decroix,
B. Tetrahedron Lett. 2001, 42, 573.
(24) (a) Meyers, A. I.; Burgess, L. E. J. Org. Chem. 1991, 56,
2294. (b) Burgess, L. E.; Meyers, A. I. J. Org. Chem. 1992,
57, 1656; and references cited therein.
(25) (a) Allin, S. M.; Duffy, L. J.; Page, P. C. B.; McKee, V.;
Edgar, M.; McKenzie, M. J.; Amat, M.; Bassas, O.; Santos,
M. M. M.; Bosch, J. Tetrahedron Lett. 2006, 47, 5713.
(b) Amat, M.; Bassas, O.; Llor, N.; Canto, M.; Perez, M.;
Molins, E.; Bosch, J. Chem. Eur. J. 2006, 12, 7872.
(26) (a) Baldwin, J. E. J. Chem. Soc., Chem. Commun. 1976,
734. (b) Baldwin, J. Tetrahedron 1982, 38, 2939.
(3) Cul, A.; Daïch, A.; Decroix, B.; Sanz, G.; Van Hijfte, L.
Tetrahedron 2004, 60, 11029.
(4) For oxa-cyclization, see: (a) Mamouni, A.; Daïch, A.;
Marchalín, Š.; Decroix, B. Heterocycles 2001, 54, 275.
(b) Sikoraiova, J.; Chihab-Eddine, A.; Marchalín, Š.; Daïch,
A. J. Heterocycl. Chem. 2002, 39, 383. (c) Pesquet, A.;
Van Hijfte, L.; Daïch, A. ARKIVOC 2010, (viii), 27.
(5) See, for example: Allin, S. M.; Gaskell, S. N.; Elsegood,
M. R. J.; Martin, W. P. J. Org. Chem. 2008, 73, 6448.
(6) (a) Aeberli, P.; Houlihan, W. J. J. Org. Chem. 1968, 33,
2402. (b) Massa, S.; De Martino, G. Farmaco 1978, 33, 271.
(7) (a) Wijnberg, J. B. P. A.; Speckamp, W. N. Tetrahedron
1978, 34, 2399. (b) Marino, J. P.; Bogdan, S.; Kimura, K.
J. Am. Chem. Soc. 1992, 114, 5566. (c) El Azab, A. S.;
Taniguchi, T.; Ogasawara, K. Org. Lett. 2000, 2, 2757.
(8) Lee, Y. S.; Kim, S. H.; Jung, S. H.; Lee, S. J.; Park, H.
Heterocycles 1994, 37, 303.
(27) Data for Compound 15a
Isolated in 35% yield as an orange solid (EtOAc–
cyclohexane = 2:3; Rf = 0.17); mp 171 °C; [a]D –212.4 (c
1.45×10–3, CH2Cl2). IR (KBr): nmax = 3019, 1686 cm–1. 1H
NMR (300 MHz, CDCl3): d = 2.02–20.7 (m, 1 H), 2.35–2.56
(m, 3 H), 3.91–4.03 (m, 1 H), 4.30 (s, 2 H), 4.81 (s, 1 H), 5.59
(s, 1 H), 7.12–7.24 (m, 5 HAr), 7.30–7.33 (m, 4 HAr), 7.44–
7.48 (t, 1 HAr, J = 6.6 Hz), 7.56 (d, 1 HAr, J = 6.0 Hz), 7.67
(d, 1 HAr, J = 7.2 Hz), 7.75 (d, 1 HAr, J = 7.2 Hz). 13C NMR
(75 MHz, CDCl3): d = 29.8, 31.5, 39.3, 51.6, 59.1, 66.6,
123.7, 124.2, 126.0, 126.4, 127.3 (2×), 127.6, 127.8, 128.8,
129.2, 129.6 (2×), 132.0, 132.5, 134.4, 135.0, 140.9, 145.0,
167.5, 175.7. MS (EI): m/z = 394 [M+]. Anal. Calcd (%) for
(9) (a) Barbosa, Y. A. O.; Hart, D. J.; Magomedov, N. A.
Tetrahedron 2006, 62, 8748. (b) Dransfield, P. J.; Dilley,
A. S.; Wang, S.; Romo, D. Tetrahedron 2006, 62, 5223.
(10) Iwamoto, O.; Koshino, H.; Hashizume, D.; Nagasawa, K.
Angew. Chem. Int. Ed. 2007, 46, 8625.
(11) (a) Yamagishi, M.; Yamada, Y.; Ozaki, K.; Date, T.;
Okamura, K.; Suzuki, M.; Matsumoto, K. J. Org. Chem.
Synlett 2011, No. 13, 1821–1826 © Thieme Stuttgart · New York

1826
J.-F. Fleury et al.
LETTER
C26H22N2O2 (394.48): C, 79.16; H, 5.62; N, 7.10. Found: C,
79.06; H, 5.52; N, 7.05.
Data for Compound 15b
HAr, J = 7.2 Hz). 13C NMR (75 MHz, CDCl3): d = 29.2, 29.8,
42.1, 50.4, 58.9, 61.8, 123.4, 124.1, 125.0, 125.7 (2×), 126.7,
127.1, 127.8 (2×), 128.6, 129.0, 131.7, 132.1, 132.3, 132.9,
133.9, 141.8, 143.7, 167.9, 176.6; MS (EI): m/z = 394 [M+].
Anal. Calcd (%) for C26H22N2O2 (394.48): C, 79.16; H, 5.62;
N, 7.10. Found: C, 78.96; H, 5.50; N, 6.93.
Isolated in 31% yield as an orange solid (EtOAc–
cyclohexane = 2:3; Rf = 0.11); mp 216 °C; [a]D –210.9 (c
0.82×10–3, CH2Cl2). IR (KBr): nmax = 3019, 1676 cm–1. 1H
NMR (300 MHz, CDCl3): d = 1.50–1.57 (m, 1 H), 2.30–2.48
(m, 2 H), 2.50–2.65 (m, 1 H), 3.53 (dd, 1 H, J = 14.1, 3.7
Hz), 4.19 (d, 1 H, J = 8.0 Hz), 4.78 (dd, 1 H, J = 14.1 Hz),
5.54–5.56 (m, 2 H), 6.23–6.29 (m, 2 HAr), 6.56 (t, 2 HAr,
J = 7.1 Hz), 6.71 (t, 1 HAr, J = 7.1 Hz), 6.84 (t, 1 HAr, J = 7.4
Hz), 6.98 (d, 1 HAr, J = 7.4 Hz), 7.13 (t, 1 HAr, J = 7.4 Hz),
7.46 (d, 1 HAr, J = 7.4 Hz), 7.55 (t, 1 HAr, J = 7.2 Hz), 7.63
(t, 1 HAr, J = 7.2 Hz), 7.81 (t, 1 HAr, J = 7.2 Hz), 7.95 (d, 1
(28) (a) See also the review in ref. 1b: (b) Maryanoff, B. E.;
McComsey, D. F. Tetrahedron Lett. 1979, 3797.
(c) Maryanoff, B. E.; McComsey, D. F.; Duhl-Emswiler,
B. A. J. Org. Chem. 1983, 48, 5062. (d) Ent, H.; de Koning,
H.; Speckamp, W. N. J. Org. Chem. 1986, 51, 1687.
(29) For the stereochemical distribution, see: Katritzky, A. R.;
Mehta, S.; He, H.-Y. J. Org. Chem. 2001, 66, 148.
Synlett 2011, No. 13, 1821–1826 © Thieme Stuttgart · New York