Conformational control of flexible molecules: Design and synthesis of novel chiral 1,5-diaza-cis-decalins

Article abstract of DOI:10.1021/jo025503x

Control of the conformational equilibria of 1,5-diaza-cis-decalins, a new class of chiral diamine ligands, has been investigated. Chiral 2,6- and 3,7-substituted derivatives of 1,5-diaza-cis-decalins were designed to stabilize the conformational form needed to chelate a lithium. These derivatives were synthesized in optically pure form starting from 1,5-diaza-cis-decalin. Due to the rigid and conformationally well-defined nature of these compounds, the potential of these compounds as chiral diamine ligands was investigated. Asymmetric lithiation- substitution reactions of N-Boc-pyrrolidine and N,N-diisopropyl-o-ethylbenzamide were performed using these ligands and up to 60% ee was obtained. For the latter substrate, results spanning a range from 32% ee (R) to 60% ee (S) were obtained (Δee = 92%) with 1,5-diaza-cis-decalin ligands differing only in the location of two methyl substituents. Unlike many other diamines that have been employed in asymmetric lithiation-substitution reactions, the limited conformational flexibility of the 1,5-diaza-cis-decalins is analogous to (-)-sparteine such that these results may permit the construction of structure-activity relationships.

Full text of DOI:10.1021/jo025503x

3072
J . Org. Chem. 2002, 67, 3072-3078
Con for m a tion a l Con tr ol of F lexible Molecu les: Design a n d
Syn th esis of Novel Ch ir a l 1,5-Dia za -cis-d eca lin s
Zhenrong Xu and Marisa C. Kozlowski*
Department of Chemistry, Roy and Diana Vagelos Laboratories, University of Pennsylvania,
Philadelphia, Pennsylvania 19104
marisa@sas.upenn.edu
Received J anuary 4, 2002
Control of the conformational equilibria of 1,5-diaza-cis-decalins, a new class of chiral diamine
ligands, has been investigated. Chiral 2,6- and 3,7-substituted derivatives of 1,5-diaza-cis-decalins
were designed to stabilize the conformational form needed to chelate a lithium. These derivatives
were synthesized in optically pure form starting from 1,5-diaza-cis-decalin. Due to the rigid and
conformationally well-defined nature of these compounds, the potential of these compounds as chiral
diamine ligands was investigated. Asymmetric lithiation-substitution reactions of N-Boc-pyrrolidine
and N,N-diisopropyl-o-ethylbenzamide were performed using these ligands and up to 60% ee was
obtained. For the latter substrate, results spanning a range from 32% ee (R) to 60% ee (S) were
obtained (∆ee ) 92%) with 1,5-diaza-cis-decalin ligands differing only in the location of two methyl
substituents. Unlike many other diamines that have been employed in asymmetric lithiation-
substitution reactions, the limited conformational flexibility of the 1,5-diaza-cis-decalins is analogous
to (-)-sparteine such that these results may permit the construction of structure-activity
relationships.
In tr od u ction
high selectivity (90-93% ee).4 However, when 2 was
employed in the asymmetric lithiation-substitution of
benzylic substrates, only moderate selectivity was achieved
(42-45% ee).3 These results stimulated us to further
examine conformational control of this flexible molecule
by means of suitably located methyl substituents. “The
conformation of a ligand is particularly important in the
development of optimum polydentate ligands. The coor-
dination of the ligand is most favorable when its ground-
state conformation resembles that of the complex.”6 Thus,
conformationally locked derivatives of 2 would facilitate
coordination of a metal since there would be no energetic
penalty associated with the chelating conformation. Such
a trait is particularly desirable in ligands for species
which form weak coordination bonds, such as alkyllithi-
ums. For the 1,5-diaza-cis-decalins, this goal cannot be
accomplished by simple substitution at the nitrogen atom
since N- or N,N′-substitution perturbs the equilibrium
toward the N-out conformation.5,7 To overcome this
problem 2,6- and 3,7-substituted-1,5-diaza-cis-decalins
were designed (Scheme 1).
Asymmetric synthesis has become a major focal point
of synthetic chemistry in the past three decades. For
metal-mediated and metal-catalyzed reactions, the design
and synthesis of appropriate chiral ligands is crucial.1
Recently, we developed a new class of chiral diamine
ligands, 1,5-diaza-cis-decalins, on the basis of computer-
aided identification of novel ligand scaffolds.2 1,5-Diaza-
cis-decalin (1) is a unique chiral diamine ligand, in which
there are two conformational isomers, N-in and N-out (eq
1). The N-in conformer has a well-defined chiral cavity
with two nitrogens that can chelate a metal.3,4 A previous
study5 showed that the parent compound 1 prefers the
N-in conformation 1a ; however, the N,N′-dimethyl deriv-
ative 2 partitions equally between the N-in and N-out
forms (2a and 2b, eq 2).
We anticipated that the methyl groups of these sub-
stituted diazadecalins would be equatorial in the N-in
form thereby “locking” the conformation. In addition,
these methyl substituents would extend the asymmetric
environment of the metal binding cavity. In this paper,
the synthesis and conformational properties of these
(2) Kozlowski, M. C.; Panda, M. J . Mol. Graph. Modell. 2002, 20,
399-409.
(3) Li, X.; Schenkel, L. B.; Kozlowski, M. C. Org. Lett. 2000, 2, 875-
878.
(4) Li, X.; Yang, J .; Kozlowski, M. C. Org. Lett. 2001, 3, 1137-1140.
(5) Santos, A. G.; Klute, W.; Torode, J .; Bo¨hm, V. P. W.; Cabrita,
E.; Runsink, J .; Hoffmann, R. W. New J . Chem. 1998, 993-997.
(6) Hoffmann, R. W. Angew. Chem., Int. Ed. Engl. 1992, 31, 1124-
1134.
In the prior studies, we have shown that 1,5-diaza-cis-
decalin (1) is the ligand of choice in catalytic enantio-
selective oxidative biaryl coupling reactions providing
(1) Seyden-Penne, J . Chiral Auxiliaries and Ligands in Asymmetric
Synthesis; J ohn Wiley & Sons: New York, 1995.
(7) Ganguly, B.; Freed, D. A.; Kozlowski, M. C. J . Org. Chem. 2001,
66, 1103-1108.
10.1021/jo025503x CCC: $22.00 © 2002 American Chemical Society
Published on Web 03/16/2002

Conformational Control of Flexible Molecules
J . Org. Chem., Vol. 67, No. 9, 2002 3073
Sch em e 1
The nature of the predominant conformation in solu-
tion of the 1,5-diaza-cis-decalins could be discerned from
the vicinal coupling constants along the molecular back-
bone.9 In the N-in conformation of 5, the nitrogen atoms
are gauche (eq 6). Thus, the angular hydrogens (H-10/
H-9) are equatorial with respect to the cyclohexyl ring
containing the adjacent H-4/H-8 protons. As a result,
small coupling constants are observed for H-10/H-9 in
this form. In the N-out conformation, the nitrogen atoms
are anti, placing the angular hydrogens axial (eq 6). As
a result, H-10/H-9 have one large (axial-axial) and one
small (axial-equatorial) coupling constant with the
adjacent H-4/H-8 protons. The 1H NMR spectrum of 5
showed H-10/H-9 as a deformed doubled doublet with
small coupling constants (J ) 2.6, 1.8 Hz; Table 1),
implying that H-10/H-9 are equatorial. In other words,
the conformational equilibrium of 5 predominantly popu-
lates the N-in conformation (eq 6).
The stereochemistry of the C-2, C-6 substituents could
also be determined by the coupling constants, particularly
between H-2/H-6 and H-3/H-7. For compound 5, H-2/H-6
participate in one large (J ) 12.3 Hz) and one small (J
) 2.9 Hz) coupling with the adjacent H-3/H-7 protons,
implying that H-2/H-6 must be axial. In other words, the
C-2/C-6 methyl substituents occupy equatorial positions.
On this basis, 5 populates mainly the N-in conformation
with the C-2, C-6 methyl groups anti to the angular
hydrogens.
It is noteworthy that the equatorial disposition of the
C-2, C-6 methyl groups substitution in 5 greatly stabilizes
the N-in conformation and even overcomes the torsional
effects of N,N′-dimethyl substitution which ordinarily
stabilizes an N-out arrangement (compare eq 1 and eq
2).7 This result confirmed that the conformational equi-
libria of 1,5-diaza-cis-decalins can be completely con-
trolled to favor the N-in conformation by suitably locating
substituent groups. On this basis, the 2,6-diphenyl-
substituted derivative 6 was designed, which should also
possess a higher conformational preference for the N-in
conformation relative to the parent compound 1.
chiral 1,5-diaza-cis-decalins are reported along with their
application to asymmetric lithiation-substitution.
Resu lts a n d Discu ssion
Syn th esis a n d Con for m a tion a l An a lysis of 2,6-
Su bstitu ted -1,5-d ia za -cis-d eca lin s. In a prior paper,8
we reported the syntheses of 2,6-dimethyl-1,5-diaza-cis-
decalins 3 and 4. On the basis of NMR analyses, isomer
3, with the C-2, C-6 methyl groups syn to the angular
hydrogens, was found to exist in the N-out conformation
(eq 3) while isomer 4, with the C-2, C-6 methyl groups
anti to the angular hydrogens, was found to exist in the
N-in conformation (eq 4). As such, 4 could act as a rigid
chelating diamine ligand due to the N-in conformational
preference. In this case, the metal would be positioned
in a well-defined extended chiral cavity.
Preparation of 6 began with the L-tartrate salt of 1,5-
diaza-cis-decalin (11) which was directly converted to the
corresponding N,N′-diBoc compound 12 by treatment
with di-tert-butyl dicarbonate in the presence of aqueous
NaOH (Scheme 2). Protected diamine 12 was conve-
niently oxidized to the corresponding bislactam 13 using
catalytic RuO2 with excess 10% aqueous NaIO4 in a two-
phase EtOAc/H2O system.10 Compound 13 was then
converted to the corresponding enol triflate 14 with N-(5-
chloro-2-pyridyl)triflimide using the procedure of Comins.11
The subsequent Pd-catalyzed Suzuki coupling reaction
The N,N′-dimethyl substituted derivative 5 was pre-
pared from 4 by Eschweiler-Clarke methylation (eq 5).
(8) Kozlowski, M. C.; Xu, Z.; Santos, A. G. Tetrahedron 2001, 57,
4537-4542.
(9) Hoffmann, R. W. Angew. Chem., Int. Ed. 2000, 39, 2054-2070.
(10) Tanaka, K.; Yoshifuji, S.; Nitta, Y. Chem. Pharm. Bull. 1988,
36, 3125-3129.
(11) Foti, C. J .; Comins, D. L. J . Org. Chem. 1995, 60, 2656-2657.

3074 J . Org. Chem., Vol. 67, No. 9, 2002
Xu and Kozlowski
Ta ble 1. 1H NMR (500 MHz, CDCl3, 23 °C) Sp ectr a l Da ta of 2,6- a n d 3,7-Su bstitu ted -1,5-d ia za -cis-d eca lin s
compound
H-10/H-9
H-2/H-6
3a
4a
5
6
7
3.09 (ddd, J ) 13.6, 5.6, 4.0 Hz, 2H)
2.68 (ddd, J ) 4.0, 2.8, 2.7 Hz, 2H)
2.04 (deformed dd, J ) 2.6, 1.8 Hz, 2H)
2.97 (dd, J ) 2.5, 2.3 Hz, 2H)
2.84 (br s, 2H)
2.85 (dqd, J ) 11.3, 6.2, 2.7 Hz, 2H)
2.64 (dqd, J ) 11.2, 6.3, 2.4 Hz, 2H)
1.98 (dqd, J ) 12.3, 6.2, 2.9 Hz, 2H)
3.75 (dd, J ) 11.2, 2.8 Hz, 2H)
2.58 (d, J ) 12.7 Hz, 2H)
2.27 (d, J ) 12.7 Hz, 2H)
8
3.06 (br d, J ) 13.1 Hz, 2H)
2.16 (d, J ) 11.5 Hz, 2H)
1.96 (dd, J ) 11.5, 1.7 Hz, 2H)
3.00 (ddd, J ) 12.3, 3.9, 2.4 Hz, 2H)
2.20 (dd, J ) 11.9, 11.3 Hz, 2H)
2.86 (ddd, J ) 10.9, 3.2, 2.4 Hz, 2H)
1.62 (dd, J ) 11.0, 10.9 Hz, 2H)
9
2.63 (dd, J ) 2.6, 2.2 Hz, 2H)
1.86 (deformed dd, J ) 2.4, 1.8 Hz, 2H)
10
a
Coupling constants determined by simulation with PPC gNMR to reproduce the second-order multiplets observed in the 1H NMR
spectra.8
Sch em e 2
insight into the conformational control of this flexible
molecule, 3,7-substituted derivatives were also examined.
Substitution of R4 appears to create a well-defined chiral
cavity in the N-in conformation (Scheme 1). However, 17
(R4 ) Me, R5 ) H) will populate mainly the N-out
conformation to minimize the 1,3-diaxial steric inter-
actions (eq 8).
Generation of a derivative in the N-in conformation
with R4 ) Me could be accomplished by geminal dimethyl
substitution of C-3 and C-7 to yield derivative 7. Com-
pound 7 was made from 13 by exhaustive methylation
of the bisenolate, followed by N-deprotection and reduc-
tion14 (Scheme 3).
of 14 with phenylboronic acid provided the corresponding
2,6-diphenyl-substituted bisenamide 15 in high yield.12
Several attempts to hydrogenate 15 using palladium
black as the catalyst were unsuccessful, even though this
catalyst was successful for a closely related transforma-
tion.13 Compound 15 was then treated with TFA which
removed the Boc groups and isomerized the resultant
bisenamine to provide bisimine 16. However, the subse-
quent reduction of 16 with DIBALH afforded a complex
mixture. The desired compound 6 was finally achieved
by the hydrogenation of bisimine 16 with 5% Rh/Al2O3
in the presence of acetic acid.
Sch em e 3
In the 1H NMR spectrum of 6, H-10/H-9 showed a
doubled doublet with small coupling constants (J ) 2.5,
2.3 Hz), implying that H-10/H-9 are equatorial (eq 7). On
the other hand, H-2/H-6 showed a doubled doublet with
a large (J ) 11.2 Hz) and a small (J ) 2.8 Hz) coupling
constant implying that H-2/H-6 are axial. From this
analysis, the N-in conformer predominates for 6 and the
C-2, C-6 phenyl groups are anti to the angular hydrogens.
The 13C NMR spectrum of 7 showed broad signals at
room-temperature implying the presence two conformers.
Interestingly, the low temperature 13C NMR at 250 K
indicated ca. 4:1 ratio of conformers with a predominance
of N-in 7a (eq 9). The N-in form 7a predominates despite
the two 1,3-interactions in the concave face of the ring
(12) Occhiato, E. G.; Trabocchi, A.; Guarna, A. Org. Lett. 2000, 2,
1241-1242.
(13) Yamaguchi, R.; Nakazono, Y.; Matsuki, M.; Hata, E.; Kawanisi,
M. Bull. Chem. Soc. J pn. 1987, 60, 215-222.
(14) Frydman, B.; Los, M.; Rapoport, H. J . Org. Chem. 1971, 36,
450-454.
Syn th esis a n d Con for m a tion a l An a lysis of 3,7-
Su bstitu ted -1,5-d ia za -cis-d eca lin s. To gain further

Conformational Control of Flexible Molecules
J . Org. Chem., Vol. 67, No. 9, 2002 3075
Ta ble 2. Exp er im en ta l Ra tios a n d Rela tive En er gies (k ca l/m ol) Ca lcu la ted w ith MM2*, Am ber *, a n d HF /6-31G* for th e
Con for m er s of 2,6- a n d 3,7-Su bstitu ted -1,5-d ia za -cis-d eca lin s
3a
3b
4a
4b
5a
5b
6a
6b
7a
80
7b
20
8a
8b
9a
9b
10a
10b
expt. ratioa
rel E (MM2*)
rel E (Amber*)
rel E (HF/6-31G*)
<10
>90
>90
<10
10.7
10.6
11.4
>90
<10
>90
<10
<10
>90
>90
<10
>90
<10
1.9
0.0
4.0
0.0
1.3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
8.3
4.3
1.1
0.0
0.0
0.0
9.3 0.0 1.0
11.6 0.0 0.3
16.3 0.0 0.7
0.0
3.4
4.2
0.0
0.0
0.0
0.0
0.0
0.0
5.9
3.9
5.6
0.0
0.0
0.0
5.0
1.2
1.9
a
Determined from the NMR data. See text.
system arising from the two axial methyl groups at C-3
and C-7. Apparently, the disfavorable interactions in 7a
(two 1,3 Me-NHR interactions) incur less of an energetic
penalty compared to those in 7b (one 1,4 H-H and four
1,3 H-CH2R interactions; see eq 9).
groups of 9 and 10 are syn to the angular hydrogens, and
these compounds predominantly populate the N-in con-
formations, 9a and 10a , respectively (eq 11 and 12). Thus,
full conformational control of 1,5-diaza-cis-decalins to
favor either the N-in or N-out forms has been achieved.
To further examine the conformational equilibria of
1,5-diaza-cis-decalins the relative energies of the con-
formers of 2,6- and 3,7-substituted derivatives were
calculated using MM2*, Amber*, and HF/6-31G* (Table
2). Generally, the calculated results support the experi-
mental observations. For the majority of the cases, the
molecular mechanics calculations were adequate; how-
ever, neither the MM2* or the Amber* calculations were
entirely reliable (see 8 and 3, respectively). These dis-
crepancies may arise from stereoelectronic interactions
in these 1,2-diamines which are not evaluated in the force
field treatments. The relevant interactions include (i)
negative hyperconjugation between the nitrogen lone-pair
and a antiperiplanar σ*-acceptor (C-H and C-C bonds)
and (ii) donation from the best σ-donor (C-H bond) to
the best σ*-acceptor (C-N bond). Previously, we had
shown that Hartree-Fock calculations are the most
reliable for the 1,5-diaza-cis-decalins.7 In line with these
results, the HF/6-31G* geometry optimized energies are
the most accurate for the entire series. On the basis of
the calculated energies, the N-in form is strongly favored
(by 5.6-16.3 kcal/mol) for the secondary amines 4, 6, and
9. The N-methyl derivatives 5 and 10 are still predis-
posed toward the N-in form, but the energy differences
(1.1 and 1.9 kcal/mol, respectively) are substantially
smaller, and some of the N-out form may be present. The
calculations indicate that N-methylation always de-
creases the stability of the N-in form relative to the
parent secondary amines (5 vs 4, 8 vs 7, 10 vs 9) due to
the introduction of torsional interactions.7 From these
results, it appears that calculations can be used to
determine if 1,5-diaza-cis-decalin derivatives are predis-
posed to one conformational form or the other.
Diamine 7 was also converted to the corresponding
N,N′-dimethyl-substituted derivative 8 by the same
Eschweiler-Clarke procedure used for 5 (Scheme 3). In
the case of 8, the H-10/H-9 protons showed a broad
doublet with a large coupling constant (J ) 13.1 Hz)
which implies that H-10/H-9 are axial and that the
undesired N-out conformation predominates (eq 10).
From the calculations (see below), diamine 7 possesses
one of the lowest proclivities for the N-in form (∆H 7a /
7b e 1 kcal/mol). N-Methylation to form 8 introduces
torsional strain between the N-Me groups and the C-4/
C-8 positions7 which offsets the marginal stabilization
of the N-in form in these tetramethyl C-3, C-7 variants.
Since 8 did not favor the desired N-in conformation, we
reevaluated the 1,5-diaza-cis-decalins and designed 3,7-
dimethyl-substituted derivatives 9 and 10.
Compounds 9 and 10 were synthesized by stereoselec-
tive alkylation15 of 13 as the key step followed by
deprotection, reduction, and methylation (Scheme 3).
NMR analyses of 9 and 10 indicated that the angular
H-10/H-9 protons are equatorial with respect to the
cyclohexyl ring containing the adjacent H-4/H-8 protons
since only small coupling constants were observed (Table
1). Thus, 9 and 10 predominantly populate the N-in
conformation. In addition, the stereochemistry of the C-3,
C-7 methyl groups of 9 and 10 was determined by the
coupling constants of H-3/H-7 with the adjacent H-2/H-6
protons. Since the H-2/H-6 axial proton signals are a
doubled doublet with large coupling constants (J ) 11.9,
11.3 Hz for 9 and J ) 11.0, 10.9 Hz for 10), H-3/H-7 must
be axial and the C-3/C-7 methyl substituents occupy
equatorial positions. On this basis, the C-3, C-7 methyl
Asym m etr ic Lith ia tion -Su bstitu tion s of N-Boc-
P yr r olidin e an d N,N-Diisopr opyl-o-eth ylben zam ide.
(15) Casimir, J . R.; Didierjean, C.; Aubry, A.; Rodriguez, M.; Briand,
J .; Guichard, G. Org. Lett. 2000, 2, 895-897.

3076 J . Org. Chem., Vol. 67, No. 9, 2002
Xu and Kozlowski
Sch em e 4
Sch em e 5
To test the effect of the modified ligands, the asymmetric
lithiation-substitution reactions of N-Boc-pyrrolidine16
and N,N-diisopropyl-o-ethylbenzamide17 were examined
(Scheme 4). Initial results from 2a gave 12% ee (S) and
45% ee (S), respectively, in these reactions. To improve
these results, a model of 2a bound to lithium (2a ‚Li,
Scheme 5) was constructed.18,19 In this model, the N-alkyl
groups, the axial H-3/H-7, and the equatorial H-2/H-6
hydrogens (circled groups in 2a ‚Li) would have the
closest contact to reactants bound to the lithium center.
Modification of the N-alkyl groups was not viable since
larger groups perturbed the equilibria of the 1,5,-diaza-
cis-decalins to the N-out form7 which does not undergo
facile chelation with alkyllithiums.20 Substitution at C-3/
C-7 was next considered since these positions are the
closest to the lithium center and may play an important
role in asymmetric induction (Scheme 5). To enhance the
stereochemical environment of the concave decalin cavity
in which the alkyllithium reagent presumably resides,
compound 8 with C-3/C-7 geminal dimethyl substitution
was designed. If the N-in conformation of 8 was predomi-
nant the axial methyl groups (circled portions of 8 in
Scheme 5) in the chiral cavity would be expected to have
a dramatic effect. However, very low selectivity (Scheme
4) was observed since 8 populated the N-out conformation
8b almost exclusively and most likely did not form a
chelated complex with the alkyllithium base.
tional preference for the N-in conformation with 10
compared to 2. Thus, 10 may act as the optimum form of
2. Alternatively, steric compression of the axial H-3/H-7
protons (circled portions of 10 in Scheme 5) in 10 may
attenuate the stereochemical environment of the 1,5-
diaza-cis-decalin cavity, resulting in the slight improve-
ment in selectivities (12 to 18% ee and 45 to 60% ee, see
Scheme 4).
Substitution at the more distal C-2/C-6 positions
(circled portions of 5 in Scheme 5) was examined next
using 5. With this compound a dramatically different
effect on the reaction selectivities was observed. Surpris-
ingly, the reactions mediated by 5 produced products with
the opposite configuration compared to 2. In an attempt
to discern the cause of this effect, the chiral environment
was analyzed by generating overlays of 2a and ent-5a
via the nitrogen atoms (Scheme 5). A topological analysis
of 2a superimposed with 5a and with ent-5a , revealed
that the shapes of 2a and ent-5a were more analogous
than those of 2a and 5a . In other words, there is a greater
degree of similarity between 2a and ent-5a relative to
2a and 5a when the overall lowest energy “absolute”
conformations (including the N-Me groups) are compared.
This analysis and the fact that the absolute stereochem-
istry is dominated by the substituents rather than the
stereochemistry of the ligand core suggests that the
periphery of ligand is much more important in defining
the stereochemical approach of the substrates in these
reactions.
Since the tetramethyl C-3/C-7 variant 8 was not viable
the dimethyl C-3/C-7 derivative 10 was examined. As
expected, 3,7-dimethyl derivative 10 gave better results
than 2. This result may arise from the higher conforma-
(16) Beak, P.; Kerrick, S. T.; Wu, S.; Chu, J . J . Am. Chem. Soc. 1994,
116, 3231-3239.
(17) Thayumanavan, S.; Basu, A.; Beak, P. J . Am. Chem. Soc. 1997,
119, 8209-8216.
(18) For reviews of asymmetric lithiation/substitution and chiral
ligand-Li complexes, see: (a) Hoppe, D.; Hense, T. Angew. Chem., Int.
Ed. Engl. 1997, 36, 2282-2316. (b) Beak, P.; Basu, A.; Gallagher, D.
J .; Park, Y. S.; Thayumanavan, S. Acc. Chem. Res. 1996, 29, 552-
560.
(19) The observed selectivity differences could also arise due to
different oligomeric species. On the basis of prior work in this field,
we begin our analysis using the simplest model possible, a monomeric
diamine-lithium adduct. For an analysis of stereoselectivity in asym-
metric deprotonation based on a monomeric sparteine‚alkyllithium
complexes, see: (a) Wiberg, K. B.; Bailey, W. F. J . Am. Chem. Soc.
2001, 123, 8231-8238. (b) Wiberg, K. B.; Bailey, W. F. Angew. Chem.,
Int. Ed. 2000, 39, 2127-2129. (c) Wiberg, K. B.; Bailey, W. F.
Tetrahedron Lett. 2000, 41, 9365-9368.
Notably, the substituent effect is different for the 2,6-
substituted derivative 5 and the 3,7-substituted deriva-
tive 10. In the case of 5, the methyl substituents not only
caused the conformational equilibrium to favor the N-in
conformation but also increased steric bulk of the C-2/
C-6 positions, providing completely different stereochem-
ical results. With 10, however, the methyl substituents
serve primarily as conformational locks. That is, 5
directly alters the stereochemical environment while 10
(20) For example, the NMR spectra of the N,N′-di-n-butyl derivative
of 1 in d8-toluene indicated the presence of the N-out form. Upon the
addition of n-BuLi in hexanes, the spectra did not indicate a change
to the N-in conformer.

Conformational Control of Flexible Molecules
J . Org. Chem., Vol. 67, No. 9, 2002 3077
residue was purified by chromatography (SiO2; 5% EtOAc/
hexane) affording 14 (1.458 g, 93%). Rf ) 0.43 (20% EtOAc/
hexane); mp 181-183 °C; [R]2D0 -121.8 (c 1.38, CHCl3); 1H
NMR (500 MHz, CDCl3) δ 5.22 (t, J ) 3.9 Hz, 2H), 4.65
(deformed t, J ) 7.0 Hz, 2H), 2.53 (ddd, J ) 18.9, 7.5, 3.8 Hz,
2H), 2.42 (ddd, J ) 18.9, 10.2, 4.0 Hz, 2H), 1.51 (s, 18H); 13C
NMR (125 MHz, CDCl3) δ 151.7, 136.8, 119.6, 117.1, 104.0,
84.2, 49.4, 27.9, 22.3; IR (film) 1727, 1683, 1213 cm-1; MS (ES)
m/z 655.1 (MNa+). Anal. Calcd for C20H26F6N2O10S2: C, 37.98;
H, 4.14; N, 4.43. Found: C, 37.67; H, 4.08; N, 4.30.
(9R ,10R )-1,5-Di-t er t -b u t oxyca r b on yl-2,6-d ip h e n yl-
1,4,5,8,9,10-h exa h yd r o-1,5-n a p h th r id in e (15). To a solution
of 14 (1.000 g, 1.58 mmol) in THF (50 mL) were added, under
a nitrogen atmosphere, phenylboronic acid (0.58 g, 4.75 mmol,
3.0 equiv), 2 M Na2CO3 (35 mL), and (Ph3P)2PdCl2 (0.111 g,
0.16 mmol, 0.1 equiv). The mixture was stirred for 5.5 h at 50
°C. Saturated brine was then added, the mixture was extracted
three times with Et2O, and the combined extracts were dried
over Na2SO4. After evaporation of the solvent, the residue was
purified by chromatography (SiO2; 5% EtOAc/hexane) affording
15 (0.749 g, 97%). Rf ) 0.40 (20% EtOAc/hexane); mp 208-
210 °C; [R]2D0 -352.2 (c 1.16, CHCl3); 1H NMR (500 MHz,
CDCl3) δ 7.24-7.32 (m, 10H), 5.26 (t, J ) 3.8 Hz, 2H), 4.93
(deformed t, J ) 7.0 Hz, 2H), 2.66 (ddd, J ) 18.8, 7.6, 3.6 Hz,
2H), 2.35 (ddd, J ) 18.9, 10.0, 4.0 Hz, 2H), 1.09 (s, 18H); 13C
NMR (125 MHz, CDCl3) δ 153.4, 141.1, 135.7, 128.0, 127.0,
125.3, 112.7, 81.0, 48.6, 27.7, 24.6; IR (film) 1699, 1646, 1344
cm-1; HRMS (ES) calcd for C30H36N2O4Na (MNa+) 511.2573,
found 511.2556. Anal. Calcd for C30H36N2O4: C, 73.74; H, 7.42;
N, 5.73. Found: C, 73.37; H, 7.36; N, 5.25.
indirectly controls the stereochemical environment. The
incorporation of conformational design into chiral ligands
will be beneficial to the development of optimum poly-
dentate ligands.21
Con clu sion
In summary, chiral 2,6- and 3,7-substituted-1,5-diaza-
cis-decalins with designed conformational preferences
were synthesized in optically pure form starting from 1,5-
diaza-cis-decalin. The nature of preferred conformations
present in solution was discerned from the coupling
constants of the angular protons H-10/H-9 with the
adjacent H-4/H-8 protons. Compounds 4, 5, 6, 9, and 10
populated predominantly the N-in conformation. Since
the N-in form of the 1,5-diaza-cis-decalins orients the two
nitrogens in a manner that allows ready chelation of a
metal, these compounds are predisposed to more ef-
fectively bind weakly coordinating species such as alkyl-
lithiums. With these compounds, the consequences of
substitution on the stereochemical course of asymmetric
lithiation-substitution reactions can be assessed since
the rigid 1,5-diaza-cis-decalin framework allows useful
structure-activity data to be collected and compared.
Further work to determine the stereochemical control
features in these reactions is underway. Investigation of
other applications of these novel chiral 1,5-diaza-cis-
decalins is also in progress.
(9R,10R)-2,6-Dip h en yl-3,4,7,8,9,10-h exa h yd r o-1,5-n a p h -
th r id in e (16). A solution of 15 (0.709 g, 1.45 mmol) and TFA
(4 mL) in CH2Cl2 (24 mL) was stirred at room temperature
for 27 h, and then the volatiles were removed in vacuo. The
residue was treated with 5% NaOH and extracted twice with
CH2Cl2. The combined organic layers were dried over K2CO3.
Filtration and concentration yielded bisimine 16 (0.421 g),
which was used in the next step without further purification.
An analytical sample was purified by chromatography (basic
Al2O3; 2% MeOH/CH2Cl2). Rf ) 0.86 (10% MeOH/CH2Cl2); mp
Exp er im en ta l Section 22
(2S,6S,9R,10R)-1,2,5,6-Tetr a m eth yl-1,5-d ia za -cis-d eca -
lin (5). A mixture of 48 (0.698 g, 4.15 mmol), formaldehyde (6
mL, 37% w/w solution), and formic acid (3 mL, 88% w/w aq
solution) was heated at 75 °C in an oil bath for 41 h, and then
ice water was added. The solution was adjusted to pH >14
with 50% aq NaOH and extracted with CH2Cl2. The extract
was dried (K2CO3) and concentrated, and the residue was
chromatographed (basic Al2O3; 3% MeOH/CH2Cl2) to afford
0.696 g (85%) of 5. Further purification could be accomplished
by Kugelrohr distillation at 60-70 °C (e1 mmHg) to give pure
5 (0.488 g, 60%) as a colorless liquid. Rf ) 0.68 (10% MeOH/
151-153 °C; [R]2D0 -99.2 (c 0.34, CHCl3); H NMR (500 MHz,
1
CDCl3) δ 7.75-7.77 (m, 4H), 7.33-7.37 (m, 6H), 4.01 (br s,
2H), 2.69-2.73 (m, 2H), 2.50-2.57 (m, 2H), 2.28-2.32 (m, 2H),
2.15-2.22 (m, 2H); 13C NMR (125 MHz, CDCl3) δ 167.4, 139.7,
129.7, 128.2, 126.1, 53.5, 25.7, 23.6; IR (film) 1632 cm-1; MS
(ES) m/z 289.2 (MH+). Anal. Calcd for C20H20N2: C, 83.30; H,
6.99; N, 9.71. Found: C, 82.84; H, 6.82; N, 9.30.
CH2Cl2); [R]20 +39.6 (c 2.43, CHCl3); 1H NMR (500 MHz,
D
CDCl3) δ 2.12 (s, 6H), 2.05 (dq, J ) 12.4, 3.1 Hz, 2H), 2.04
(deformed dd, J ) 2.6, 1.8 Hz, 2H), 1.98 (dqd, J ) 12.3, 6.2,
2.9 Hz, 2H), 1.70 (tdd, J ) 13.1, 11.0, 2.9 Hz, 2H), 1.32 (tt, J
) 13.9, 3.3 Hz, 2H), 1.25 (dq, J ) 12.6, 3.1 Hz, 2H), 1.12 (d, J
) 6.2 Hz, 6H); 13C NMR (125 MHz, CDCl3) δ 63.8, 60.6, 37.8,
29.7, 28.7, 21.4; IR (film) 2963, 2772 cm-1; HRMS (ES) calcd
for C12H24N2Na (MNa+) 219.1837, found 219.1844.
(2R,6R,9R,10R)-2,6-Dip h en yl-1,5-d ia za -cis-d eca lin (6).
A mixture of 16 and 5% Rh/Al2O3 (0.147 g) in MeOH (9 mL)
and HOAc (1 mL) was placed under 40 psi H2 at room
temperature. After 48 h, the mixture was diluted with CH2-
Cl2 and washed with 30% NaOH. The aqueous phase was
extracted twice with CH2Cl2. The combined organic layers were
dried over K2CO3. After the solvent was evaporated under
reduced pressure, the residue was chromatographed (SiO2; 2%
MeOH/CHCl3) to give 43% of 6 (0.180 g) in two steps from 15.
(9R,10R)-1,5-Di-ter t-bu toxyca r bon yl-2,6-bis[(tr iflu or o-
methylsulfonyl)oxy]-1,4,5,8,9,10-hexahydro-1,5-naphthridine
(14). To a solution of 138 (0.919 g, 2.49 mmol) in THF (20 mL)
at -78 °C was added 7.5 mL (7.5 mmol, 3.0 equiv) of a 1.0 M
LiHMDS solution by syringe pump at 0.2 mL/min rate. After
an additional 6 h, N-(5-chloro-2-pyridyl)triflimide23 (3.02 g, 3
equiv) in THF was added. The resulting mixture was stirred
overnight, with the temperature slowly rising to ambient. The
reaction mixture was quenched with saturated brine. The
organic layer was diluted with Et2O, and the aqueous phase
was extracted twice with Et2O. The combined organic extracts
were dried over MgSO4. After evaporation of the solvent, the
Rf ) 0.42 (20% MeOH/CHCl3); [R]20 -1.34 (c 1.19, CHCl3); 1H
D
NMR (500 MHz, CDCl3) δ 7.42 (d, J ) 7.1 Hz, 4H), 7.35 (t, J
) 7.6 Hz, 4H), 7.26 (t, J ) 7.3 Hz, 2H), 3.75 (dd, J ) 11.2, 2.8
Hz, 2H), 2.97 (dd, J ) 2.5, 2.3 Hz, 2H), 1.89-1.92 (m, 4H),
1.74-1.88 (m, 2H), 1.74 (br s, 2H), 1.63 (dq, J ) 12.8, 3.2 Hz,
2H); 13C NMR (125 MHz, CDCl3) δ 145.2, 128.4, 127.0, 126.5,
61.9, 53.3, 31.6, 29.0; IR (film) 3424, 2922, 698 cm-1; HRMS
(CI) calcd for C20H24N2 (M+) 292.1939, found 292.1945.
(3R,7R,9R,10R)-N,N′-Di-ter t-bu toxyca r bon yl-2,6-d ioxo-
3,7-d im eth yl-1,5-d ia za -cis-d eca lin (20). To a solution of 138
(1.605 g, 4.35 mmol) in THF (60 mL) at -78 °C was added 9.2
mL (9.2 mmol, 2.1 equiv) of a 1.0 M LiHMDS solution by
syringe pump at 0.12 mL/min rate. After an additional 6.5 h,
MeI (2.2 mL, 8 equiv) was added. The resulting mixture was
stirred overnight, with the temperature slowly rising to
ambient. The reaction mixture was quenched with saturated
brine and extracted three times with Et2O. The combined
organic extracts were dried (Na2SO4) and concentrated, and
(21) For some examples, see: (a) Wang, X.; Erickson, S. D.; Iimori,
T.; Still, W. C. J . Am. Chem. Soc. 1992, 114, 4128-4137. (b) Desper,
J . M.; Gellman, S. H.; Wolf, J r., R. E.; Cooper, S. R. J . Am. Chem. Soc.
1991, 113, 8663-8671. (c) Villacorta, G. M.; Rao, C. P.; Lippard, S. J .
J . Am. Chem. Soc. 1988, 110, 3175-3182. See also refs 6 and 9.
(22) See Supporting Information for the General Procedures section.
(23) (a) Comins, D. L.; Dehghani, A.; Foti, C. J .; J oseph, S. P. Org.
Synth. 1997, 74, 77-83. (b) Comins, D. L.; Dehghani, A. Tetrahedron
Lett. 1992, 33, 6299-6302.

3078 J . Org. Chem., Vol. 67, No. 9, 2002
Xu and Kozlowski
the residue was chromatographed (SiO2; 20% EtOAc/hexane)
to afford 20 (0.842 g, 49%). Rf ) 0.39 (50% EtOAc/hexane);
mp 92-93 °C; [R]2D0 -95.3 (c 1.60, CHCl3); 1H NMR (500 MHz,
CDCl3) δ 4.55-4.59 (m, 2H), 2.68-2.75 (m, 2H), 2.12-2.18 (m,
2H), 1.96 (ddd, J ) 14.0, 7.4, 4.8 Hz, 2H), 1.53 (s, 18H), 1.30
(d, J ) 7.2 Hz, 6H); 13C NMR (125 MHz, CDCl3) δ 173.4, 152.5,
83.9, 51.8, 35.3, 32.4, 27.9, 17.5; IR (film) 1769, 1720, 1251,
1151 cm-1; MS (ES) m/z 419.2 (MNa+). Anal. Calcd for
mixture was stirred for 40 min at -78 °C and then added to
a precooled solution of N,N-diisopropyl-o-ethylbenzamide (0.200
g, 0.857 mmol, 1.0 equiv) in MTBE (13 mL)/pentane (13 mL)
at -78 °C. The resulting purple reaction mixture was stirred
for 6 h at -78 °C, and then 0.3 mL (4.0 equiv) of allyl bromide
was added. The resulting mixture was stirred overnight, with
the temperature slowly rising to ambient. The reaction mixture
was quenched with MeOH (0.5 mL). After 0.5 h, the reaction
mixture was diluted with 5% H3PO4 and extracted three times
with Et2O. The combined Et2O extracts were washed with
brine, dried (Na2SO4), and concentrated in vacuo. The residue
was purified by chromatography (SiO2; 10% EtOAc/hexane)
to give alkylated product (0.169 g, 72%). The analytical data
were identical to those reported by Beak.17 The enantiomeric
purity was determined by HPLC using a Pirkle Covalent (R,R)-
â-GEM 1 column (25 cm × 4.6 mm) with 1% i-PrOH/hexane
at 1 mL/min: tR (R) ) 10.9 min, tR (S) ) 11.7 min.
C
20H32N2O6: C, 60.59; H, 8.13; N, 7.06. Found: C, 60.39; H,
8.38; N, 7.02.
(3R,7R,9R,10R)-2,6-Dioxo-3,7-d im et h yl-1,5-d ia za -cis-
d eca lin (21). A solution of 20 (0.842 g, 2.12 mmol) and TFA
(5 mL) in CH2Cl2 (30 mL) was stirred at room temperature
for 16 h, and then the volatiles were removed in vacuo. MeOH
was added to the residue and then was removed under reduced
pressure affording 21 (0.420 g, 100%) as a white solid. Rf ) 0
1
(EtOAc); mp > 280 °C; H NMR (500 MHz, CDCl3) δ 5.60 (br
s, 2H), 3.90 (s, 2H), 2.54-2.62 (m, 2H), 2.00 (ddd, J ) 14.4,
5.4, 2.2 Hz, 2H), 1.78 (deformed t, J ) 13.7 Hz, 2H), 1.25 (d,
J ) 7.1 Hz, 6H); 13C NMR (125 MHz, CDCl3) δ 174.5, 49.5,
35.0, 29.5, 15.9; IR (film) 3188, 1651, 1622 cm-1. Anal. Calcd
for C10H16N2O2: C, 61.20; H, 8.22; N, 14.27. Found: C, 61.03;
H, 8.53; N, 14.08.
The combined aqueous layers were adjusted to pH > 14 with
30% NaOH and then extracted three times with CH2Cl2, dried
(K2CO3), and concentrated in vacuo. The residue was chro-
matographed (basic Al2O3; 3% MeOH/CH2Cl2) to afford 0.237
g (90%) of 10, identical spectral data and rotation with the
sample above.
(3R,7R,9R,10R)-3,7-Dim eth yl-1,5-d ia za -cis-d eca lin (9).
A mixture of 21 (0.417 g, 2.12 mmol) and LiAlH4 (2 g, 20 equiv)
in THF (60 mL) was heated at reflux for 53 h, and then the
mixture was cooled in ice-water. Excess Na2SO4‚10H2O was
added. The resulting mixture was diluted with Et2O and
filtered through filter paper, and the solid was washed with
Et2O. The filtrate was concentrated under reduced pressure
yielding 9 (0.360 g, 100%) as a waxy solid. Rf ) 0.56 (Al2O3;
10% MeOH/CH2Cl2); [R]20 +31.4 (c 2.53, CHCl3); 1H NMR
(500 MHz, CDCl3) δ 3.00 D(ddd, J ) 12.3, 3.9, 2.4 Hz, 2H), 2.63
(dd, J ) 2.6, 2.2 Hz, 2H), 2.20 (dd, J ) 11.9, 11.3 Hz, 2H),
1.81 (br s, 2H), 1.69-1.78 (m, 2H), 1.23-1.30 (m, 2H), 0.77 (d,
J ) 6.4 Hz, 6H); 13C NMR (125 MHz, CDCl3) δ 55.0, 53.9, 39.9,
26.3, 19.4; IR (film) 3285, 2949, 2924, 1651, 1456 cm-1; HRMS
(ES) calcd for C10H21N2 (MH+) 169.1705, found 169.1706.
(3R,7R,9R,10R)-1,3,5,7-Tetr a m eth yl-1,5-d ia za -cis-d eca -
lin (10). A mixture of 9 (0.357 g, 2.12 mmol), formaldehyde
(10 mL, 37% w/w aq solution), and formic acid (5 mL, 88%
w/w aq solution) was heated at 85 °C in an oil bath for 36 h.
The cooled solution was adjusted to pH >14 with 30% aq
NaOH and extracted three times with CH2Cl2. The extract was
dried (K2CO3) and concentrated, and the residue was chro-
matographed (basic Al2O3; 2% MeOH/CH2Cl2) to afford 10
(0.341 g, 82%) as a solid. Rf ) 0.76 (10% MeOH/CH2Cl2); mp
52-63 °C (hexane); [R]D20 -26.7 (c 1.98, CHCl3); 1H NMR (500
MHz, CDCl3) δ 2.86 (ddd, J ) 10.9, 3.2, 2.4 Hz, 2H), 2.15 (s,
6H), 2.02-2.07 (m, 4H), 1.86 (deformed dd, J ) 2.4, 1.8 Hz,
2H), 1.62 (dd, J ) 11.0, 10.9 Hz, 2H), 0.92 (td, J ) 13.6, 2.6
Hz, 2H), 0.79 (d, J ) 6.4 Hz, 6H); 13C NMR (125 MHz, CDCl3)
δ 65.8, 63.3, 42.7, 37.0, 25.5, 19.6; IR (film) 2944, 2776, 1453
cm-1; HRMS (ES) calcd for C12H25N2 (MH+) 197.2018, found
197.2026.
Typ ica l P r oced u r e for Asym m etr ic Lith ia tion -Su b-
stitu tion of N-Boc-P yr r olid in e. To a solution of 10 (0.289
g, 1.47 mmol, 1.3 equiv) in Et2O (10 mL) at -78 °C was added
s-BuLi (1.14 mL, 1.29 M in cyclohexane, 1.47 mmol, 1.3 equiv).
The reaction mixture was stirred for 25 min, and then a
solution of N-Boc-pyrrolidine (0.188 g, 1.10 mmol, 1.0 equiv)
in Et2O (10 mL) was added by cannula. The reaction was
allowed to stir for 6 h, TMSCl (0.3 mL, 2 equiv) was added,
and the reaction was allowed to warm slowly to room-
temperature overnight. The reaction was quenched by addition
of 5% H3PO4, extracted with Et2O, dried (Na2SO4), and
concentrated in vacuo. The residue was chromatographed
(SiO2; 2-3% EtOAc/hexane) to afford 0.079 g (30%) of product.
The analytical data were identical to those reported by Beak.16
The enantiomeric purity was determined by GC using a chiral
capillary Supelco â-Dex 120 column (30 m × 0.25 mm) at 95
°C and 30 psi: tR (S) ) 43.0 min, tR (R) ) 44.2 min.
Diamine 10 was recovered in 88% yield from the aqueous
phase as described above.
Ackn owledgm en t. We thank Dr. Manoranjan Panda
for performing the energy calculations of the 1,5-diaza-
cis-decalin conformers. Financial support was provided
by the National Institutes of Health (GM59945), Merck
Research Laboratories, DuPont. Acknowledgment is
made to the donors of the Petroleum Research Fund,
administered by the American Chemical Society, for
partial support of this research.
Su p p or tin g In for m a tion Ava ila ble: Detailed experi-
mental procedures and characterization data of compounds 18,
Typ ica l P r oced u r e for Asym m etr ic Lith ia tion -Su b-
st it u t ion of N,N-Diisop r op yl-o-et h ylb en za m id e. To a
solution of 10 (0.262 g, 1.34 mmol, 1.5 equiv) in MTBE (13
mL)/ pentane (13 mL) at -78 °C was added s-BuLi (1.05 mL,
1.23 M in cyclohexane, 1.30 mmol, 1.5 equiv). The reaction
1
19, 7, and 8 are provided as well as copies of the H NMR and
13C NMR spectra for compounds 3-10. This material is
available free of charge via the Internet at http://pubs.acs.org.
J O025503X