Mild synthesis of enantiomerically pure imidazo-[1,2-a]azepines mediated by Yb(OTf)3

Article abstract of DOI:10.1016/S0040-4020(01)01078-X

The reaction of various chiral 2,2′-diaryldialdehydes with achiral and chiral 1,2-diamines in the presence of Lewis acids to give imidazo[1,2-a]azepines was investigated. Best results were achieved with Yb(OTf)₃; the reaction outcome is strongly dependent upon the geometric features of both reactants. Kinetic resolution of rac-2,2′-dinaphthyldialdehyde with (R,R)-1,2-diphenyl-1,2-diamminoethane (up to 92% e.e.) was achieved.

Full text of DOI:10.1016/S0040-4020(01)01078-X

TETRAHEDRON
Pergamon
Tetrahedron 57 -2001) 10357±10363
Mild synthesis of enantiomerically pure imidazo-[1,2-a]azepines
mediated by YbꢀOTf)3
Rita Annunziata, Maurizio Benaglia and Laura Raimondip
Á
Dipartimento di Chimica Organica e Industriale and Centro CNR, Universita di Milano, Dipartimento di Chimica Organica e Industriale
and Centro CNR Studio, Sintesi Stereochimica Speciali Sistemi Organici, via Camillo Golgi 19, I-20133 Milano, Italy
Received 26 June 2001; revised 4 October 2001; accepted 1 November 2001
AbstractÐThe reaction of various chiral 2,20-diaryldialdehydes with achiral and chiral 1,2-diamines in the presence of Lewis acids to give
imidazo[1,2-a]azepines was investigated. Best results were achieved with Yb-OTf)3; the reaction outcome is strongly dependent upon the
geometric features of both reactants. Kinetic resolution of rac-2,20-dinaphthyldialdehyde with -R,R)-1,2-diphenyl-1,2-diamminoethane -up
to 92% e.e.) was achieved. q 2001 Elsevier Science Ltd. All rights reserved.
1. Introduction
in the presence of Yb-OTf)3 unexpectedly led to the diaster-
eoisomerically pure adduct -R,R,R)-3 -93% chemical yield;
Scheme 1).
²
The biological activity of molecules containing a modi®ed
azepinic ring is intensively tested, both in vitro and in
vivo, against various diseases -e.g. cancer, AIDS-related
dementia, Alzheimer's disease).1,2 In the course of our
studies toward the synthesis of chiral 1,2-diamines,3 we
observed that treatment of -R)-6,60-dimethyldiphenyl-2,20-
dialdehyde 1 with -R,R)-1,2-diphenyl-1,2-diaminoethane 2
Actually, this reaction outcome is not completely unprece-
dented: inspection of the literature con®rmed that in acidic
conditions the condensation reaction between diphenyl-
2,20-dialdehydes with 1,2-diamines led to the synthesis of
imidazo[1,2-a]azepines via formation of an hydroimidazole,
Scheme 1. Synthesis of -R,R,R)-3 in the presence of Yb-OTf)3.
Keywords: imidazo[1,2-a]azepines; ytterbium tri¯ate; imine; kinetic resolution.
Corresponding author; e-mail: lauramaria.raimondi@unimi.it
p
²
The formation of -R,R,R)-3 was ®rst observed when the reaction was performed in THF and in the presence of SmI2 together with Yb-OTf)3. However, Sm-II)
species did not interfere with the reaction, and only led to traces of a dimeric product 4 -Scheme 1), derived from 3 via hydrogen abstraction and dimerization
of the benzylic radical.
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020-01)01078-X

10358
R. Annunziata et al. / Tetrahedron 57 .2001) 10357±10363
Table 1. Reaction of dialdehyde rac-5 with different 1,2-diamines under
acidic conditions. Structures of reactants and adducts are given in Fig. 1 and
Scheme 1
Table 2. Reaction of chiral dialdehydes -R)-1 and -R)-13 with different
diamines in the presence of Yb-OTf)3. Structures of reactants and adducts
are given in Figs. 1 and 2 and Scheme 1
Entry
1,2-Diamine T -8C)
Acid
Product
rac-7
rac-7
±
rac-7
Yield%a
Entry
Dialdehyde
Diamine
T -8C)
Product
Yield%a
1
2
3
4
5
6
7
8
6
25
40
25
25
25
25
25
40
Yb-OTf)3
Yb-OTf)3
Yb-OTf)3
TiCl4
CF3COOH
Yb-OTf)3
Yb-OTf)3
Yb-OTf)3
47
40
±
46
27
5
1
2
3
4
5
6
-R)-1
-R)-1
-R)-1
-R)-1
-R)-13
-R)-13
-R,R)-2
-R,R)-9
-R,R)-9
6
-R,R)-2
-S,S)-2
25
25
40
25
25
25
-R,R,R)-3
-R,R,R)-11
-R,R,R)-11
-R)-12
-R,R,R)-14
±
93
23b
25b
45
41
±
6
6
6
6
b
rac-7
-R,R)-2
-R,R)-9
-R,R)-9
-Rp,R,R)-8
-Rp,R,R)-10
-Rp,R,R)-10
12
18
a
Yields refer to products puri®ed by ¯ash chromatography.
Yield was estimated by 1H NMR -the product decomposed on silica gel).
b
a
Yields refer to products puri®ed by ¯ash chromatography.
0.2 M equiv.
b
Surprisingly enough, from the reaction of rac-5 with 1,2-
diaminobenzene 6 the expected product 7 was isolated in
47% yield only -entry 1). No improvements were observed
upon performing the reaction in re¯uxing CH2Cl2 -entry 2),
while the use of catalytic Yb-OTf)3 did not afford
reasonable quantities of the adduct -only unreacted 5 and
6 were isolatedÐentry 3). We turned our attention to the
use of other acidic species, in an attempt to increase the
chemical yields; however, the results were disappointing.
Only in the presence of TiCl4 the reaction proceeded with
decent, but not improved, yield -entry 4). The use of strong
protic species such as CF3COOH -1 M equiv.) led to the
formation of 7, but only in 27% yield -entry 5).
subsequent intramolecular nucleophilic addition of one
of the imidazole nitrogens to the aldehydic carbon, loss of
water and 1,3-hydride shift.4,5 However, the use of
Yb-OTf)3 as Lewis acid allows the reaction to be carried
on under milder conditions -12 h at room temperature
³
instead of 24 h in re¯uxing acetic acid). Moreover,
chemical yields seemed to be excellent. We thus undertook
more detailed studies, in order to verify the applicability of
this improved methodology to a general synthesis of
imidazo[1,2-a]azepines.
Furthermore, diamine -R,R)-2, which gave excellent results
in the reaction with dialdehyde -R)-1 -Scheme 1), was
almost unreactive with rac-5: only traces of the correspond-
ing imidazo[1,2-a]azepine 8 were isolated -entry 6). The use
of -R,R)-1,2-diaminocyclohexane 9 was also disappointing:
adduct 10 was isolated in only 12% yield when the reaction
was performed at room temperature, and in slightly higher
yields -18%) at 408C -entries 7 and 8).
2. Results and discussion
Racemic diphenyl-2,20-dialdehyde 5 was initially chosen as
a substrate, and its reaction with various 1,2-diamines in the
presence of different acidic species was examined. All reac-
tions were run in CH2Cl2 as solvent for 12 h in the presence
of 1 M equiv. of Yb-OTf)3 under a nitrogen atmosphere.
Signi®cant results are collected in Table 1. Structures of
the reactants and adducts are shown in Fig. 1 and Scheme
1. Adducts were fully characterized by spectroscopic tech-
niques, as discussed later.
Based on these results, we decided to re-examine the
reaction of -R)-1 with different diamines; the results are
Figure 1. Structures of rac-5, 1,2-diamines 6, -R,R)-9 and adducts 7, 8 and 10 -Table 1).
³
Yb-OTf)3 was chosen because of its recognized ability to activate imines toward nucleophilic attack at carbon. See Ref. 3 and references therein.

R. Annunziata et al. / Tetrahedron 57 .2001) 10357±10363
10359
1,3-diamine) a complex mixture of products was isolated,
with no formation of the expected imidazo[1,2-a]azepine.
The reaction of chiral, enantiomerically pure -R)-2,20-
dinaphthyldialdehyde 13 with -R,R)-2 led to the expected
enantiomerically pure imidazo[1,2-a]azepine -R,R,R)-14 as
the unique product in 41% chemical yield -entry 5). On the
other hand, no reaction occurred between -R)-13 and the
mismatching diamine -S,S)-2 -entry 6), thus demonstrating
the existence of both the matching -R)-13 plus -R,R)-2 and
the mismatching -R)-13 plus -S,S)-2 couples.
1
2.1. Structural elucidation by H and 13C NMR
spectroscopy
A complete spectroscopic investigation was performed to
establish the structure of all reaction products depicted in
Figs. 1 and 2, and in Scheme 1. A concerted use of several
1
gradient-enhanced experiments such as 2D COSY, H/13C
2D HMQC -Heteronuclear Multiple-Quantum Coherence)6
and 1H/13C 2D HMBC -Heteronuclear Multiple-Bond
Correlation)6±8 provided the sequence-speci®c assignment
of all compounds allowing the structural elucidation of these
multi-ring aromatic derivatives. The 1H and 13C NMR
chemical shifts values are reported in Section 4.
Figure 2. Structures of dialdehyde -R)-13 and adducts 11, 12 and 14
-Table 2).
collected in Table 2, structures of reactants and adducts in
Figs. 1 and 2, and Scheme 1.
From the multiplicity and the 1H/1H connectivities found in
2D COSY spectra, we were able to distinguish the H8±H10
and H80 ±H100 series of the protons for compounds 3, 11,
12, as H8±H11 and H80 ±H110 protons for 7, from the other
aromatic hydrogens H14±H6 and H17±H19 for 3, or
H12±H15, for 12, but it was impossible to assign each series
of these three -four) protons to the right ring -for the proton
and carbon numbering, see Fig. 4). For this purpose we
performed the heteronuclear direct- and multiple-bond
We were pleased to ®nd that the reaction of -R)-1 with
-R,R)-2 in the presence of Yb-OTf)3 allowed for the forma-
tion of -R,R,R)-3 in excellent chemical yield -entry 1). The
use of different diamines led again to disappointing results:
with -R,R)-9 only the corresponding monoimine 11 was
formed, at room temperature as well as in re¯uxing
CH2Cl2 -entries 2 and 3); only one diastereoisomer was
isolated. In the reaction with 1,2-diaminobenzene 6, adduct
12 was isolated in 45% yield -entry 4). The reaction of -R)-1
with ¯exible 1,5-diaminopentane and 1,6-diaminohexane
only led to the formation of polymeric imines; on the
other hand, with 1,8-diaminonaphthalene -a rigid, planar
1
correlations experiments: the H/13C HMQC gave for all
1
protons the direct attachment to carbons, while the H/13C
HMBC allowed the complete assignments of all protons and
quaternary carbons. The correlation of C3 with H8 and C5
with H80 for all compounds, and of C3 with H7 and C5 with
Figure 3. Dihedral angles, N±N and CvO intramolecular distances for some of the diamines and dialdehydes considered in the present work.

10360
R. Annunziata et al. / Tetrahedron 57 .2001) 10357±10363
Ê
-2.8 vs 2.9 A) and an almost planar arrangement -11 vs 08)
in both cases -Fig. 3). A 1,2-relationshipbetween the two
amino groups evidently avoids strain in the formation of a
fused imidazolinic ring, the ®rst step in the proposed reac-
tion path.3,4 Some structural data for the diamines and the
dialdehydes used in this work, as evaluated in vacuo by
molecular mechanic techniques, are shown in Fig. 3.¶
Similar considerations apply for aliphatic and alicyclic
1,2-diamines. When reacted with -R)-1, ¯exible -R,R)-2
can better be accommodated on the forming imidazolinic
ring than -R,R)-9 -Table 2); in the ®rst case, the imi-
dazo[1,2-a]azepinic ring is formed in excellent yields,
while in the latter, the equilibria are shifted toward the
formation of the corresponding monoimine. Reasonably
enough, 1,5- and 1,6-diamines led only to dimeric imines,
while formation of intramolecular bis-imines or imi-
dazo[1,2-a]azepine is completely disfavored.
Also the differences between the reactivity of dialdehydes 1,
5 and 13 can be ascribed to their structural differences -Fig.
3). The two ortho methyl groups lock the aryl rings of
dialdehyde 1 in an almost orthogonal conformation
-2878), while in the less constrained dialdehyde 5 the
dihedral angle closes upto 2678. This decreases the
Ê
distance between the two carbonyls from 3.8 to 3.2 A.
Differences between the conformations of 2,20-dinaphthyl-
dialdehyde 13 and of 6,60-dimethyl-2,20-diphenyldialde-
hyde 1 are even smaller: both compounds show an almost
identical dihedral angle between the two aryl groups, the
steric hyndrance of an ortho-fused phenyl ring being
comparable to the one of an ortho-methyl residue. The
two carbonyls, however, are slightly closer in 2,20-
Scheme 2. Kinetic resolution of rac-13 with -R,R)-2.
0
Ê
dinaphthyldialdehyde 13 -3.6 A) than in 6,6 -dimethyl-
0
Ê
2,2 -diphenyldialdehyde 1 -3.8 A). The reaction mechanism
greatly ampli®es these apparently insigni®cant differences
-Tables 1 and 2), thus revealing a strong dependence of
the reaction path upon geometrical constrains of both
reactants.
H6 in the case of 7 and 12, was the key to assign the
hydrogens of the diaryl moiety and, as a consequence, the
other set of aromatic protons. This was supported by
the correlation of C6 with H14, in compounds 3, 7, 12,
14, and of C7 with H17 -3, 14) or H13 -7, 12). Moreover,
the C20 carbon showed both long-range couplings with H5
and H8, which allowed us to distinguish C2 from C20, as the
correlation with H8 or H80 distinguish C1 from C10. The
methylene protons H5 gave an AB system, with a geminal
coupling constant of about 14 Hz, in 3, 7, 12, 14.
Our results, together with literature data,3,4 clearly indicate
the limits of this reaction; however, good to excellent results
can be obtained with chiral, enantiomerically pure, match-
ing reactants, as in the case of -R)-1 and -R)-13 with -R,R)-2
-Table 2, entries 1 and 5). To further test this point, we
undertook an investigation of the feasibility of a kinetic
resolution of rac-13:11 clearly, only -R)-13 would react in
the presence of diamine -R,R)-2, thus allowing for a maxi-
mum 50% yield. Indeed, in the reaction of rac-13 with
-R,R)-2, we isolated the expected imidazo[1,2-a]azepine
-R,R,R)-14 in 13% yield; unreacted -S)-13 -15% yield)
showed a 92% optical purity -Scheme 2). The multi-step
nature of the reaction mechanism, with formation of
several intermediates, can account for the low yield of
the process. Attempts to increase the chemical yield by
raising the reaction temperature, however, led to the forma-
tion of a diastereoisomeric mixture of -R,R,R)- and -S,R,R)-
14 in a 2:1 ratio and 21% yield -Scheme 2), while in the
presence of only 0.5 M equiv. of -R,R)-2 the reaction did not
occur.
2.2. Discussion
Our results allow some interesting considerations: ®rst of
all, the structure of the diamine plays the major role in
determining the feasibility of the reaction. With planar,
aromatic 1,2-diamines, such as 1,2-diaminobenzene 6,
similar chemical yields are observed with both 2,20-di-
phenyldialdehyde 5 and 6,60-dimethyl-2,20-diphenyldialde-
hyde 1.§ Obviously, a 1,2-planar arrangement of the two
reacting nitrogens guarantees a smooth reaction path.
Other planar diamines, such as 1,8-diaminonaphthalene,
did not react as expected despite a similar N±N distance
§
Actually, in Refs. 4 and 5, authors employed 4,5-diaminopyrimidine and
2,3-diaminonaphthalene; the results were extremely similar to those
obtained by us with 1,2-diaminobenzene 6.
¶
Molecular mechanic calculations were run on a SGI O2 workstation, with
the MacroModel 5.5 package9 and the MM2p force ®eld.10

R. Annunziata et al. / Tetrahedron 57 .2001) 10357±10363
10361
Figure 4. Numbering of imidazo[1,2-a]azepinic rings for NMR assignments.
3. Conclusion
pressure. Dry CH2Cl2 was distilled from CaH2 and kept
under a nitrogen atmosphere.
The scope and limitations of a mild Lewis acid catalyzed
synthesis of imidazo[1,2-a]azepines have been studied, and
the structure of the products was elucidated by NMR tech-
niques. The outcome of the process and the stereochemistry
of the adducts are completely dominated by geometrical
parameters of the reactants -as elucidated by molecular
mechanics): when the reaction occurs, it is completely
stereo- and enantio-selective. Matching and mismatching
pairs play a major role when dealing with chiral reagents,
to the point that only one enantiomer may lead to the hetero-
cyclization reaction.
Compounds -R,R)-2, -S,S)-2, 6, -R,R)-9 were commercially
available. -R)-1,12 rac-5,13 rac-14 and -R)-1315 were prepared
following known procedures.
4.1. General procedure for the synthesis of imidazo-
[1,2-a]azepines
To a stirred solution -1023 M) of the desired dialdehyde in
dry CH2Cl2, the required amount of Lewis or protic acid
-Tables 1 and 2) and 1 mol of the diamine were added;
the reaction was allowed to stir for 12 h either at room
temperature -258C) or at 408C, in the presence of MgSO4
Ê
or molecular sieves -4 A) as drying agents. The reaction was
4. Experimental
monitored via TLC -silica gel). The reaction mixture was
®ltered, and after removal of solvent the organic residue
was puri®ed by ¯ash chromatography. Chemical yields are
collected in Tables 1 and 2. For the numbering system used
in the description of the NMR spectra, see Fig. 4.k
CHN analyses were performed on a Perkin±Elmer 240
1
instrument. H and 13C NMR spectra were recorded on a
Bruker AM300 at 300.133 and 75.47 MHz, respectively, in
CDCl3 as solvent. 1H chemical shifts are reported in d rela-
tive to TMS; H±H coupling constants and 13C chemical
shifts are reported in Hz. IR Spectra were recorded in
CH2Cl2 solution on a Perkin±Elmer 1725X FTIR spectro-
meter. Mass spectra were performed on an analytical VG
7070 EQ instrument. Optical activities were measured
with a Perkin±Elmer 241 polarimeter. Melting points are
uncorrected.
4.1.1. Compound ꢀR,R,R)-3. The compound was a waxy
compound; [Found: C, 86.96; H, 6.31; N, 6.73. C30H26N2
20
requires C, 86.92; H, 6.32; N 6.76%]; Rf -Et2O) 0.4; [a]D
ˆ
170.0 -c 0.3, CH2Cl2); nmax -CH2Cl2) 3055, 2930, 2360,
2341, 1735, 1602, 1454, 1263 cm21; dH -300.133 MHz,
CDCl3) 8.12 -1H, d, Jˆ7.5 Hz, H-8), 7.51 -1H, d, Jˆ
7.5 Hz, H-10), 7.46 -1H, t, Jˆ7.5 Hz, H-9), 7.37±7.33
Silica gel -230±400 mesh) was used for ¯ash chroma-
tography. Organic extracts were dried over Na2SO4 and
®ltered before removal of the solvent under reduced
k
We were not able to achieve a satisfactory puri®cation of adducts 8 and
10; thus, no complete spectroscopic and chemical characterization was
possible in these two cases.

10362
R. Annunziata et al. / Tetrahedron 57 .2001) 10357±10363
-4H, m, H-16, H-18, H-19), 7.28 -2H, d, Jˆ7.0 Hz, H-17),
7.24 -2H, t, Jˆ7.0 Hz, H-15), 7.20 -1H, d, Jˆ7.5 Hz,
H-100), 6.97 -2H, d, Jˆ7.0 Hz, H-14), 6.92 -1H, t, Jˆ
7.5 Hz, H-90), 6.17 -1H, d, Jˆ7.5 Hz, H-80), 5.15 -1H, d,
Jˆ6.8 Hz, H-7), 4.77 -1H, d, Jˆ6.8 Hz, H-6), 4.31 -1H, B
part of an AB system, Jˆ14.0 Hz, H-5B), 3.73 -1H, A part of
an AB system, Jˆ14.0 Hz, H-5A), 2.19 -3H, s, CH3), 2.14
-3H, s, CH3); dC -75.47 MHz, CDCl3) 165.0 -C-3), 140.0
-C-20), 138.5 -C-1, C-2, C-11), 138.2 -C-110), 135.0 -C-10),
134.0 -C-10), 130.5 -C-100), 129.4 -C-14, C-16, C-18,
C-19), 128.7 -C-8), 128.5 -C-9), 128.1 -C-14), 128.0
-C-90), 126.7 -C-17), 125.0 -C-80), 77.0 -C-6), 73.0 -C-7),
48.5 -C-5), 20.5, 2.19 -CH3); m/z -FAB1): 415 -MH1).
H-130, H-16), 7.00 -2H, d, Jˆ7.1 Hz, H-14), 6.68 -1H, d,
Jˆ8.2 Hz, H-80), 5.07 -1H, d, Jˆ8.0 Hz, H-7), 4.68 -1H, d,
Jˆ8.0 Hz, H-6), 4.53 -1H, B part of an AB system, Jˆ
14.5 Hz, H-5B), 3.84 -1H, A part of an AB system, Jˆ
14.5 Hz, H-5A); dC -75.47 MHz, CDCl3) 165.0 -C-3),
139.0 -C-20), 135.0 -C-2), 134.5 -C-1, C-a), 133.4 -C-a0),
133.2 -C-b0), 132.7 -C-10, C-b), 129.05 -C-18, C-19), 129.0
-C-9, C-15, C-16), 128.9 -C-90), 128.6 -C-10), 128.5
-C-100), 128.1 -C-130, C-14), 127.6 -C-13), 127.3 -C-11),
127.1 -C-17), 126.5 -C-120), 126.3 -C-110, C-12), 125.9
-C-8), 125.8 -C-80), 77.6 -C-6, C-7), 48.3 -C-5); m/z
-FAB1): 487 -MH1). From the kinetic resolution experi-
ment, dialdehyde -S)-13 was recovered with 92% optical
purity: [a]Dˆ27.7 vs 28.4 -EtOH).15
4.1.2. Compound rac-7. The compound was a white solid,
mp150 8C; [Found: C, 85.05; H, 4.98; N, 9.97. C20H14N2
requires C, 85.08; H, 5.00; N, 9.92]; Rf -50% hexane/
Et2O) 0.19; nmax -CH2Cl2) 3057, 2982, 2361, 1736, 1692,
1459, 1449, 1266, 1245 cm21; dH -300.133 MHz, CDCl3)
8.31 -1H, dd, Jˆ7.1, 2.2 Hz, H-8), 7.86 -1H, dd, Jˆ7.4,
1.5 Hz, H-12), 7.66 -1H, dd, Jˆ7.1, 2.2 Hz, H-9), 7.62
-3H, m, H-10, H-11, H-110), 7.57 -1H, dd, Jˆ7.4, 1.5 Hz,
H-15), 7.48 -1H, dd, Jˆ7.5, 1.5 Hz, H-80), 7.45 -1H, dt,
Jˆ7.5, 1.5 Hz, H-100), 7.40 -1H, t, Jˆ7.5, 1.5 Hz, H-90),
7.35 -1H, dt, Jˆ7.5, 1.5 Hz, H-14), 7.27 -1H, dt, Jˆ7.5,
1.5 Hz, H-13), 5.23 -1H, B part of an AB system, Jˆ
14.0 Hz, H-5B), 5.00 -1H, A part of an AB system, Jˆ
14.0 Hz, H-5A); dC -75.47 MHz, CDCl3) 152.1 -C-3),
143.5 -C-7), 139.2 -C-10), 138.0 -C-1), 136.3 -C-2, C-20),
134.5 -C-6), 130.8 -C-110), 130.6 -C-8), 130.2 -C-10), 130.0
-C-11), 129.0 -C-100), 128.4 -C-9), 128.3 -C-90), 127.5
-C-80), 122.5 -C-14), 122.2 -C-13), 120.0 -C-12), 108.7
-C-15), 46.8 -C-5).
4.1.5. Compound monoimine ꢀR,R,R)-11. The compound
was not chromatographed and was isolated as an oil;
[Found: C, 79.07; H, 7.81; N, 8.34. C22H26N2O requires C,
79.00; H, 7.84; N, 8.38]; nmax -CH2Cl2) 3520, 3505, 3055,
2930, 2360, 2342, 1695, 1645, 1602, 1263 cm21; dH
-300.133 MHz, CDCl3) 9.63 -1H, s, H-3), 8.06 -1H, d, Jˆ
7.7 Hz, H-80), 7.91 -1H, dd, Jˆ7.7, 0.9 Hz, H-8), 7.76 -1H,
s, H-5), 7.61 -1H, d, Jˆ7.7 Hz, H-10), 7.52 -1H, t, Jˆ
7.7 Hz, H-52), 7.35 -2H, m, H-90, H-100), 3.23 -1H, m,
H-7), 2.83 -1H, m, H-6), 2.17 -2H, m, H-12), 1.67 -2H, m,
H-15), 1.33 -4H, m, H-13, H-14), 2.00 -3H, s, CH3), 1.88
-3H, s, CH3).
Acknowledgements
We thank Professors Mauro Cinquini and Franco Cozzi
for helpful discussions, and Dr Sara Nemfardi for her
collaboration.
4.1.3. Compound ꢀR)-12. The compound was isolated as a
waxy compound; [Found: C, 85.17; H, 5.83; N, 9.00.
C22H18N2 requires C, 85.13; H, 5.85; N, 9.02]; Rf -Et2O)
0.45; [a]Dˆ132.7 -c 0.2, CH2Cl2); nmax -CH2Cl2) 3057,
2935, 2361, 1735, 1658, 1456, 1265 cm21; dH -300.133
MHz, CDCl3) 8.01 -1H, dd, Jˆ7.1, 2.2 Hz, H-8), 7.81
-1H, dd, Jˆ7.0, 1.8 Hz, H-12), 7.50±7.49 -2H, m, H-9,
H-15), 7.45 -1H, dd, Jˆ7.1, 1.8 Hz, H-10), 7.31±7.25
-3H, m, H-80, H-13, H-14), 7.19 -2H, m, H-90, H-100),
5.08 -1H, B part of an AB system, Jˆ14.2 Hz, H-5B),
4.88 -1H, A part of an AB system, Jˆ14.2 Hz, H-5A),
2.30 -3H, s, CH3), 2.10 -3H, s, CH3); dC -75.47 MHz,
CDCl3) 152.0 -C-3), 143.5 -C-7), 138.8 -C-20, C-110),
137.5 -C-2, C-11), 136.3 -C-1, C-10), 134.5 -C-6), 131.3
-C-10), 130.5 -C-100), 128.0 -C-8), 127.9 -C-9), 127.8
-C-90), 124.3 -C-80), 122.4 -C-14), 122.1 -C-13), 120.1
-C-12), 108.6 -C-15), 47.2 -C-5), 20.2 -CH3), 20.0 -CH3).
References
1. Guzikowski, A. P.; Cai, S. X.; Espitia, S. A.; Hawkinson, J. E.;
Huettner, J. E.; Nogales, D. F.; Tran, M.; Woodward, R. M.;
Weber, E.; Keana, J. F. W. J. Med. Chem. 1996, 39, 4643±
4653 and references therein.
2. Link, A.; Kunick, C. J. Med. Chem. 1998, 41, 1299±1305 and
references therein.
3. Annunziata, R.; Benaglia, M.; Cinquini, M.; Cozzi, F.;
Raimondi, L. Tetrahedron Lett. 1998, 39, 3333±3336.
4. Rabinovitz, M.; Agranat, I.; Shaw, W.-C. J. Org. Chem. 1981,
46, 3018±3020.
5. Agranat, I.; Rabinovitz, M.; Tang, C.-P.; Shaw, W.-C. J. Org.
Chem. 1982, 47, 1578±1579.
6. Hurd, R. E.; John, B. K. J. Magn. Reson. 1991, 91, 648.
7. Ruiz-Cabello, J.; Vuister, G. W.; Moonen, C. T. W.; van
Gelder, P.; Cohen, J. S.; van Zijl, P. C. M. J. Magn. Reson.
1992, 100, 282.
4.1.4. Compound ꢀR,R,R)-14. The compound was a thick
oil; [Found: C, 88.91; H, 5.36; N, 5.73. C36H26N2 requires C,
88.86; H, 5.39; N, 5.75]; Rf -Et2O) 0.25; [a]Dˆ119.7 -c 1.3,
CH2Cl2); nmax -CH2Cl2) 3055, 2986, 2360, 2307, 1734,
1374, 1266, 1246 cm21; dH -300.133 MHz, CDCl3) 8.42
-1H, d, Jˆ8.5 Hz, H-8), 8.09 -1H, d, Jˆ8.5 Hz, H-9), 8.01
-1H, t, Jˆ8.2 Hz, H-10), 7.88 -1H, d, Jˆ8.2 Hz, H-100),
7.63 -1H, d, Jˆ8.2 Hz, H-90), 7.53 -1H, dt, Jˆ8.2, 1.5 Hz,
H-11), 7.44 -1H, dt, Jˆ8.2, 1.0 Hz, H-110), 7.37 -5H, m,
H-17, H-18, H-19), 7.30 -1H, t, Jˆ7.1 Hz, H-16), 7.22 -1H,
dt, Jˆ8.2, 1.0 Hz, H-12), 7.18±7.14 -5H, m, H-120, H-13,
8. Wilker, W.; Leibfritz, D.; Kerssebaum, R.; Bermel, W. Magn.
Reson. Chem. 1993, 31, 287.
9. Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liskamp, R.;
Lipton, M.; Cau®eld, C.; Chang, G.; Hendrickson, T.; Still,
W. C. J. Comput. Chem. 1990, 11, 440±467.
10. Burkert, U.; Allinger, N. L. Molecular Mechanics; ACS
Monograph, Vol. 177; American Chemical Society:
Washington, DC, 1982.

R. Annunziata et al. / Tetrahedron 57 .2001) 10357±10363
10363
11. Keith, J. M.; Larrow, J. F.; Jacobsen, E. N. Adv. Synth. Catal.
2001, 343, 5±26 and references therein.
H.; Kitahiro, K.; Matsumoto, H. Chem. Pharm. Bull. 1977, 25,
3312±3323.
14. Brunner, H.; Goldbrummer, J. Chem. Ber. 1989, 2005±2010.
15. Annunziata, R.; Benaglia, M.; Cinquini, M.; Cozzi, F.;
Woods, C. R.; Siegel, J. S. Eur. J. Org. Chem. 2001, 173±180.
12. -a) Andrus, M. B.; Asgari, D.; Sclafani, J. A. J. Org. Chem.
1997, 62, 9365±9368. -b) Tichy, M.; Holanova, J.; Zavada, J.
Tetrahedron: Asymmetry 1998, 9, 3497±3504.
13. Kobayashi, S.; Kihiara, M.; Shizu, S.; Katayama, S.; Ikeda,