A theoretical and experimental study of the asymmetric addition of dialkylzinc to N-(diphenylphosphinoyl)benzalimine

Article abstract of DOI:10.1002/(SICI)1521-3765(19990604)5:6<1692::AID-CHEM1692>3.0.CO;2-M

The mechanism of the enantioselective addition of diethylzinc to N-(diphenylphosphinoyl)benzalimine with catalysis by bicyclic 2-azanorbornyl-3-methanols was studied by quantum chemical calculations. The mechanism proved to differ from that of the addition of diethylzinc to aldehydes and also from an earlier proposed mechanism. The results of the calculations were used to identify several factors responsible for the selectivity. The theoretical evaluation was performed in connection with an experimental study of the effects of introducing an additional stereocenter in the ligand. An efficient route to both diastereomers of new bicyclic 2-azanorbornyl-3-methanols with an additional chiral center (the secondary alcohol group) is also presented. In the best case, an enantiomeric excess of up to 97% was obtained with these new ligands.

Full text of DOI:10.1002/(SICI)1521-3765(19990604)5:6<1692::AID-CHEM1692>3.0.CO;2-M

FULL PAPER
A Theoretical and Experimental Study of the Asymmetric Addition of
Dialkylzinc to N-(Diphenylphosphinoyl)benzalimine
Peter Brandt,[b] Christian Hedberg,[a] Klaus Lawonn,[a] Pedro Pinho,[a] and
Pher G. Andersson*[a]
Abstract: The mechanism of the enan-
tioselective addition of diethylzinc to N-
(diphenylphosphinoyl)benzalimine with
catalysis by bicyclic 2-azanorbornyl-3-
methanols was studied by quantum
chemical calculations. The mechanism
proved to differ from that of the addi-
tion of diethylzinc to aldehydes and also
from an earlier proposed mechanism.
The results of the calculations were used
to identify several factors responsible
for the selectivity. The theoretical eval-
uation was performed in connection
with an experimental study of the effects
of introducing an additional stereocen-
ter in the ligand. An efficient route to
both diastereomers of new bicyclic
2-azanorbornyl-3-methanols with an ad-
ditional chiral center (the secondary
alcohol group) is also presented. In the
best case, an enantiomeric excess of up
to 97% was obtained with these new
ligands.
Keywords: ab initio calculations ´
amino alcohols ´ asymmetric catal-
ysis ´ nucleophilic additions ´ zinc
amino alcohols possessing the 2-azanorbornyl skeleton 1 and
their applications in asymmetric catalysis.[5]
Introduction
Carbon ± carbon bond formation plays an important role in
synthetic organic chemistry, and addition of organometallic
reagents to carbonyl groups is one of the most popular
methods for achieving this purpose.[1, 2] In modern organic
chemistry many of these addition reactions were performed
with high asymmetric induction by using chiral auxiliaries or
ligands. Among the most successful classes of ligands are
optically active b-amino alcohols, which are easily synthesized
from the corresponding a-amino acids and are extremely
useful in many asymmetric reactions.[3] In particular, ligands
derived from proline have been successfully used in a variety
of transformations,[4] a fact that inspired us to search for even
more powerful ligands based on the proline backbone. We are
especially interested in the synthesis of chiral, nonnatural b-
Such bicyclic ligands with a primary alcohol group (R1 ˆ
R2 ˆ H) are effective in the ruthenium-catalyzed asymmetric
transfer hydrogenation of ketones,[6] and their tertiary ana-
logues (R1 ˆ H, R2 ˆ Ar) are good promotors in the asym-
metric reduction of ketones via the corresponding in-situ-
generated oxazaborolidines.[7] High enantioselectivities in the
nucleophilic addition of dialkylzinc reagents to N-(diphenyl-
phosphinoyl)imines[8] (Scheme 1) were also attained in the
presence of stoichiometric or catalytic amounts of the N-
substituted 2-azanorbornyl-3-methanols (R1 ˆ Bn; R2 ˆ H,
Me, iPr, or Ph).[9]
[a] P. G. Andersson, C. Hedberg, K. Lawonn, P. Pinho
Department of Organic Chemistry
Uppsala University, Box 531
To obtain more knowledge about the factors governing the
selectivity, we performed a quantum chemical investigation of
S-751 21 Uppsala (Sweden)
Fax: (‡46)18-512524
E-mail:phera@akka.kemi.uu.se
[b] P. Brandt
Department of Medicinal Chemistry
Royal Danish School of Pharmacy
Universitetsparken 2
DK-2100 Copenhagen (Denmark)
Supporting information for this article is available on the WWW under
http://www/wiley-vch.de/home/chemistry/ or from the author.
Scheme 1. a) Et2Zn (3 equiv), chiral amino alcohol (1 equiv), toluene, 08C
to room temperature.
1692
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0506-1692 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 6

1692±1699
the reaction mechanism of the latter reaction. By combining
new experimental results with the results of the calculations,
we were able to reach several conclusions about the selectivity
process. We also report an efficient route to both diaster-
eomers of bicyclic 2-azanorbornyl-3-methanol compounds
with an additional stereocenter (2), that is, a secondary
alcohol group, and their ability to mediate the formation of
chiral amines in the addition of dialkylzinc to prochiral imines
(Scheme 1).
Results and Discussion
Reaction Mechanism
Successful rationalizations of enantioselectivity depend on
comprehensive knowledge about the reaction mechanism.
The mechanism of amino alcohol catalyzed addition of
dialkylzinc reagents to aldehydes was recently investigated
by quantum chemical methods.[10] This reaction take places via
a [2.2.0] bicyclic transition state (TS) analogous to A in
Figure 1. With this structure as a starting point, we evaluated
five types of TS for the addition of diethylzinc to N-
(diphenylphosphinoyl)benzalimine (Figure 1).
Figure 1. Five different types of transition states (TS) evaluated for the
addition of diethylzinc to N-(diphenylphosphinoyl) benzalimine.
from precoordinated N-(dimethylphosphinoyl)benzalimine
As shown in Table 1a [2.2.0] bicyclic TS A, as in the case of
the analogous aldehyde reaction, can be excluded from the
potential energy surface of the imine reaction. The two
reactions evidently differ in reaction mechanism. Structure B,
a ring-opened analogue of A, could not be located for our
large model system and instead optimized to structure A. For
a smaller model system this TS was about 3 kcalmol 1 higher
in energy than D (see Supporting Information). Structure C,
which contains an even more tightly bound substrate,
optimized to structure D and could not be located in any of
the calculations. A transition state such as D was earlier
invoked to explain the stereochemical outcome of the
reaction.[11a] As shown in Table 1 an analysis based on this
TS indeed results in a correct prediction of the enantioselec-
tivity. This TS is also considerably lower in energy than A.
However, a TS in which the phosphinoyl oxygen atom
coordinates to the Zn center of the catalyst is even lower in
energy. In addition, the calculated energy difference between
the two lowest energy TSs that lead to opposite enantiomers
of the product fits accurately to the experimental value of the
enantioselectivity (91% ee).[9] The energy barrier calculated
1
for this TS (Table 1, entry 5) is 12.7 kcalmol
at the
1
HF/3-21G level and 14.9 kcalmol for the application of
B3PW91 to the HF geometry. This barrier is reasonably
low, and the conclusion must be that the reaction occurs
via a TS of type E. The driving forces calculated similarly from
1
the same precoordinated complex are 54.8 kcalmol and
1
40.1 kcalmol , respectively. One reason for this TS being
lower in energy than D is probably that it avoids major
steric repulsion between the ligand and the aryl group of the
imine. Calculations on a smaller model system with the
sterically unhindered N-(dimethylphosphinoyl)formimine
support this conclusion (see Supporting Information).[2]
Origin of the selectivity: In rationalizing the enantioselectiv-
ity of the addition of diethylzinc to N-(diphenylphosphinoyl)-
benzalimine, 16 possible diastereomeric transition states of
type E must be considered. To perform this extensive analysis,
the TSs were categorized according to four different geo-
metrical parameters, each capable of blocking a specific set of
diastereomeric routes.
Table 1. Relative transition state energies [kcalmol 1] for the addition of dimethylzinc to N-(dimethylphosphinoyl)benzalimine with ligand 1 (R1 ˆ Me,
R2 ˆ H) as catalyst.
Entry
TS type
enantiomer
Product
at Zn1
Configuration
eq/ax[a]
HF/3-21G
B3PW91/[b]
//HF/3 ± 21G
B3PW91/[b]
//B3PW91/[c]
1
2
3
4
5
6
7
8
9
A
D
D
D
E
E
E
E
E
S
S
S
R
S
R
R
S
R
R
S
R
R
S
34.8
12.7
21.2
18.0
0.0
2.2
7.7
1.1
2.5
22.4
12.2
16.5
15.2
0.0
2.7
9.5
1.8
2.6
11.7
0.0
eq
eq
ax
eq
eq
S
R
R
R
1.4
[a] Orientation of the aryl substituent of the imine in the TS of type E. [b] The basis set was 6 ± 311 ‡ G* for Zn and 6 ± 31G* for P, C, N, O, and H. [c] The
basis set was 6 ± 311 ‡ G for Zn and 6 ± 31G for P, C, N, O, and H.
Chem. Eur. J. 1999, 5, No. 6
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0506-1693 $ 17.50+.50/0
1693

P. G. Andersson et al.
FULL PAPER
General ligand effects: Coordination of the nitrogen atom of
the ligand to zinc (Zn1, Figure 1) requires an R configuration
at this position that places the alcohol group and the
coordinated zinc center in a syn arrangement. This lowers
the number of energetically viable TSs to eight.
The second diastereo-differentiating component in the
reaction can be referred to as equatorial/axial selectivity, in
which the geometrical distinction concerns the orientation of
the aryl substituent of the imine in the TS. Figure 2 illustrates
Substituent effects: As we reported before,[9] the use of a
tertiary alcohol group (1; R1 ˆ Bn, R2 ˆ Ph) led to a poor ee
(16%), and the best result (91% ee) for this bicyclic amino
alcohol was obtained with a primary alcohol group (1; R1 ˆ
Bn, R2 ˆ H). The best result so far was achieved with an
aziridino alcohol (94% ee).[11a]
A potential site for introducing new substituents is the
position a to the alcohol in the 2-azanorbornyl-3-methanol
ligand. A series of such ligands 5 ± 8 was synthesized with the
intention of studying the effects of substituents on the
enantioselectivity of the reaction, and the results are listed
in Table 2. Ligand 5 gave the highest enantiomeric excess
Table 2. Addition of diethylzinc to N-(diphenylphosphinoyl)benzalimine.
Entry
Chiral ligand
Yield [%][a]
ee [%][b]
Abs. config.[c]
1
2
3
4
5
6
7
8
70
68
59
52
97
93
79
71
S
S
S
S
[a] Yield of isolated product after flash chromatography (silica gel;
pentane/acetone). [b] Determined by HPLC analysis on a chiral column
(ChiralCel OD-H). [c] Determined as described in ref. [11].
reported so far (97% ee) and thus makes this reaction an even
more powerful tool for the synthesis of chiral amines. All of
these reactions were performed with a stoichiometric amount
of the chiral ligand under the conditions previously report-
ed.[9,11]
Figure 2. The two lowest transition states found for the addition of
dimethylzinc to N-(dimethylphosphinoyl)benzalimine (entries 5 and 8 in
Table 1), optimized at the B3PW91 level with the basis sets 6 ± 311 ‡ G for
Zn and 6 ± 31G for other atoms. Hydrogen atoms omitted for clarity.
To identify the factors responsible for the increase in ee with
ligand 5, we performed two large calculations on this system
(Table 3). The calculated energies correlate well with the
two TSs with an equatorial-type orientation of the aryl group.
This effect proved to be very strong and was estimated to be
about 7 kcalmol in favor of an equatorial configuration
1
Table 3. Substituent effects on energies [kcalmol 1] and selected geo-
metrical parameters.[a]
(Table 1, entries 7 and 8). The number of TSs to be included
in the analysis of the enantioselectivity is thus reduced to
four.
Chiral Product
B3PW91/[b]
Out-of-plane angle Dihedral angle
ligand enantiomer //HF/3 ± 21G (Zn1-O-Zn2-Ca)
D(C-Zn2-N-C)
The preference for an R configuration at Zn1 is due to the
steric requirements of the ligand in the back-folded (S)-Zn1-
1
5
1
5
(S)
(S)
(R)
(R)
0.0
0.0
1.8
2.8
1408
1688
1538
1558
88
48
98
1
TS. This effect can be estimated to be around 3 kcalmol
128
(Table 1, entries 5 and 6) and thus may leave room for
competitive pathways.
[a] The comparison refers to the lowest (S)- and the lowest (R)-TS in
Table 1. [b] The basis set was 6-311 ‡ G* for Zn and 6-31G* for P, C, N, O,
and H.
Finally, the face selectivity at the imine determines the
enantioselectivity of the reaction. The calculations correctly
predict a preferential formation of the S product, and the
energy difference calculated by B3PW91 on the HF geometry
experimental results in Table 2 and thus allow a deeper
analysis of the selectivity-determining process. The main
structural differences between TS structures with unsubsti-
tuted and substituted ligands concern the out-of-plane tor-
sional angle Zn1-O-Zn2-Ca (a to the alcohol) and the dihedral
angle C-Zn2-N-C in the four-membered ring of the TS. These
parameters are affected oppositely by the substituent in the
1
is 1.8 kcalmol (Table 1). Experimentally, this energy differ-
1
ence was estimated at about 2 kcalmol for a measured ee of
91%.[9] The main difference between the lowest (R)- and the
lowest (S)-TS concern the orientation of the four-membered
ring (Zn2-C-C-N), where an exo orientation is favored
(Figure 2).
1694
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0506-1694 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 6

Asymmetric Addition
1692±1699
two diastereomeric TSs. Perhaps the most important of these
effects is the deviation from planarity of the four-membered
ring. Comparing, for example, entries 6 and 9 in Table 1, for
which the main differences are the exo/endo relationship of
the four-membered ring (endo for entry 6) this effect seems to
be efficiently compensated by the difference in planarity. The
TS in entry 6 is almost completely planar whereas that of
entry 9 is puckered (C-Zn2-N-C 318).
erated in situ and treated at low temperature with 12 to give
the desired secondary alcohols with good selectivity (85:15)
and in up to almost quantitative yield. The diastereomeric
mixture could be easily puri-
fied by flash chromatography
and the major isomers isolated
as white solids. The stereose-
lective introduction of the sec-
ond chiral center into the bicy-
clic compound was thus ach-
Ligand synthesis
ieved. The high selectivity of
the reaction is due to prefer-
The synthetic approach to the key intermediate 12 is
outlined in Scheme 2. Compound 9 was obtained by a
diastereoselective aza-Diels ± Alder reaction between cyclo-
Figure 3. Preferential
ap-
ential approach of the nucleo-
phile from the less hindered
side of the nonchelated[15]
a-amino aldehyde (Figure 3).
Hydrogenolysis of 13 was
proach of the nucleophile from
the less hindered side of the
aldehyde.
first attempted with a catalytic amount of Pd/C,[18] but this
did not yield the desired product. However, treatment of the
same compound under 20 atm H2 in presence of catalytic
amounts of Pearlmanꢁs catalyst (Pd(OH)2/C) gave the free
amino alcohols 14 as white solids in excellent yield.
Having isolated the diastereomers 14, we thought the
Mitsunobu reaction[19] would be suitable for inverting the
absolute configuration at the stereocenter of the secondary
alcohol. However, the reaction did not take place, probably
due to steric hindrance. Instead we developed another short
and straightforward route to the diastereomers starting from
prochiral bicyclic ketones (Scheme 3).
Scheme 2. i) H2 (1 atm), Pd/C (10 wt%), EtOH, room temperature.
ii) LiAlH4, THF, room temperature. iii) Swern oxidation. iv) RMgX/
CeCl3, THF, 788C. v) H2 (300 psi), Pd(OH)2/C (20 wt%), EtOH, room
temperature. vi) Cl2CO, NEt3, THF, 08C. vii) BnCl, K2CO3, MeCN, room
temperature.
pentadiene and the imine derived from ethyl glyoxylate and
(S)-1-phenylethylamine.[13] After purification by flash chro-
matography the major exo cycloadduct was hydrogenated
under an atmosphere of molecular hydrogen in the presence
of Pd/C (10 wt%) to yield the N-protected amino ester 10.
Reduction with LiAlH4 and subsequent oxidation of 11 under
Swern conditions led to the a-amino aldehyde 12 as a yellow
oil in excellent yield.[14] Treatment of 12 with Grignard
reagents afforded adduct 13 in low yield, even when reverse
addition was used.[15] Addition of Grignard reagents to
carbonyl groups can be promoted by addition of anhydrous
cerium chloride, which presumably forms highly oxophilic
organocerium reagents.[16, 17] These intermediates were gen-
Scheme 3. i) (S)-1-phenylethylamine, CH2Cl2, 08C; TFA, BF3 ´ OEt2, cy-
clopentadiene, CH2Cl2, 788C. ii) H2 (1 atm), Pd/C (10 wt%), MeOH,
K2CO3, room temperature. iii) LiAlH4, THF, 788C. iv) H2 (300 psi),
Pd(OH)2/C (20 wt%), EtOH, room temperature. v) Cl2CO, NEt3, THF,
08C. vi) BnCl, K2CO3, MeCN, room temperature.
The prochiral ketones were obtained by a diastereoselec-
tive aza-Diels ± Alder reaction analogous to that described for
the synthesis of the ester 9. Treatment of cyclopentadiene with
the imines prepared from (S)-1-phenylethylamine and phe-
nylglyoxal or pyruvic aldehyde resulted in the cycloadducts
Chem. Eur. J. 1999, 5, No. 6
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0506-1695 $ 17.50+.50/0
1695

P. G. Andersson et al.
FULL PAPER
16. Hydrogenation of 16 under an atmosphere of hydrogen in
presence of 10% Pd/C gave the a-amino ketones 17.
Reduction of these intermediates with NaBH4 was unsuc-
cessful (20% yield after 2d at 208C), but with LiAlH4 the
corresponding alcohols were obtained in good yields and high
selectivity (80:20). With diisobutylaluminum hydride (DI-
BAL) a similar yield was obtained but the selectivity was
slightly lower (70:30). The N-protected b-amino alcohols 18
were subsequently hydrogenolyzed and benzylated under the
conditions employed for the diastereomers 13.[18]
Scheme 5. Determination of the absolute stereochemistry.[20]
Conclusions
Determination of stereochemistry
On the basis of quantum chemical investigations with the
B3PW91[21] hybrid functional we conclude that the reaction
mechanism, both for the reaction catalyzed by bicyclic
2-azanorbornyl-3-methanols and for the reaction catalyzed
by aziridinoalcohols,[11] occurs via a TS of type E (Figure 1).
This finding enabled the factors that govern the enantiose-
lectivity of the reaction to be analyzed, and four elements of
the selectivity process were identified. We improved the
enantioselectivity of the reaction up to 97% ee by introducing
an additional chiral center in the ligand. Work is now in
progress to uncover the mechanistic details of the catalytic
cycle. In addition, a TS molecular mechanics model is
currently being developed to enable faster evaluations of
substituent effects and rapid evaluation of new ligands.
Relative stereochemistry: Treatment of 14 and 19 with
phosgene furnished the corresponding N,O-carbamates 15
and 20 (Schemes 2 and 3). The relative configuration of these
compounds was determined by NOE difference spectroscopy
(Figure 4).
Experimental Section
Methods of calculation: All calculations were performed with the
Gaussian 94 program.[22] Geometry optimizations were performed by HF
calculations with the 3-21G basis set. For some selected transition states,
geometry optimizations were performed with B3PW91, a density functional
type of calculation with a hybrid functional, together with the 6-31G basis
set for all atoms except for zinc, for which 6-311 ‡ G was used. The final
energies were determined with B3PW91 and the 6 ± 31G* basis set for all
atoms except for zinc (6-311 ‡ G*). The model used in the calculations was
very close to the real system, although some truncations were necessary to
reduce the size of the system. As the aim of the calculations was to
rationalize the stereochemical out-
Figure 4. NOE difference spectroscopy of the N,O-carbamates 15 and 20
confirming the relative stereochemistry of the a-substituents.
Absolute stereochemistry: Further clues to the absolute
stereochemistry of the compounds 19 was obtained by
performing a Swern oxidation on the 2-azanorbornyl-3-
methanols 13 to give the b-amino ketones 17 (Scheme 4).
Since all spectroscopic data were in complete agreement with
that observed for the ketones derived from the usual aza-
Diels ± Alder reaction followed by hydrogenation, it could be
concluded that the two compounds have the same absolute
configuration.
come of the reaction, all stereocenters
were left intact. The model system is
depicted in Figure 5. The imine was
considered to exist exclusively in a
trans configuration.
General methods: For general exper-
imental information, see ref. [23]. All
reactions were performed under ni-
Figure 5. Model system used in
trogen or argon. Before use the re-
the calculations.
action vessels were evacuated, heat-
ed, and flushed with inert gas. All
commercially available reagents were employed as supplied, except for
cyclopentadiene, which was freshly distilled prior to use. Flash chromatog-
raphy was performed on silica gel (Matrex 60A, 37 ± 70 mm). Deactivated
silica gel was prepared by elution of the column with 5% Et3N in pentane
until the eluate was alkaline (pH paper). TLC was performed on precoated
plates (SIL G-60 UV254, Macherey-Nagel). Deactivated silica gel was
obtained by eluting TLC plates with 5% Et3N in pentane and drying before
Scheme 4.
The chiral induction by the (R)-1-phenylethylamine in the
aza-Diels ± Alder reaction between cyclopentadiene and the
imine derived from ethyl glyoxylate and (R)-1-phenylethyl-
amine was assigned by means of X-ray crystallography
(Scheme 5).[20] Since 16a and 16b have the same absolute
stereochemistry as the 2-azanorbornyl-3-methanols 13, an
assignment of the absolute stereochemistry could be made for
all compounds derived from these intermediates.
1
applying the sample. H and 13C NMR spectra were recorded on a Varian
Gemini 200 spectrometer, a Varian XL 300 spectrometer, or a Varian Unity
400 spectrometer at 258C; the reference for 1H chemical shifts was residual
CHCl3 (d ˆ 7.26), and that for 13C chemical shifts the 13C signal of CDCl3
(d ˆ 77.0). Mass spectra were recorded at 70 eV with a Finnigan MAT GCQ
instrument by direct inlet. IR spectra were recorded with a Perkin Elmer
1760 FTIR spectrometer and a Perkin Elmer 1600 series FTIR spectrom-
1696
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0506-1696 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 6

Asymmetric Addition
1692±1699
eter. HPLC analysis of diethylzinc addition products was carried out on a
chiral column (ChiralCel OD-H), with a UV detector and a flow rate of
0.5 mLmin 1 of hexane/isopropyl alcohol (95/5). The retention times were
24.7 min for the R isomer and 36.9 min for the S isomer. Et2Zn was
purchased from Aldrich Co. Imine 1 was prepared according to a literature
procedure.[9b]
(10 mL) was added and the product was extracted with CH2Cl2 (3 Â
10 mL). After drying over Na2SO4 the solvent was evaporated and the
residue was purified by flash chromatography. Yield: 78%; Rf ˆ 0.20
(EtOAc/pentane 1/1); [a]D25
ˆ
21.0 (c ˆ1.00 in CHCl3); IR (CHCl3): nÄ ˆ
1
3055, 2987, 2306, 1424, 1271 cm
;
1H NMR: d ˆ 7.40 ± 7.20 (m, 5H), 3.73
3
(brs, 1H), 3.57 (q, J(H,H) ˆ 6.4 Hz, 1H), 3.45 (brs, 1H), 2.31 (brs, 1H),
1.95 ± 2.15 (m, 2H), 1.90 (d, 3J(H,H) ˆ 9.8 Hz, 1H), 1.80 ± 1.55 (m, 2H), 1.42
(d, 3J(H,H) ˆ 6.4 Hz, 5H), 1.35 ± 1.15 (m, 3H), 0.75(d, 3J(H,H) ˆ 6.0 Hz,
3H), 0.53 (d, 3J(H,H) ˆ 6.0 Hz, 3H); 13C NMR: d ˆ 145.5, 128.3, 127.6,
127.4, 75.6, 70.1, 60.8, 58.2, 37.1, 36.4, 30.5, 30.0, 22.5, 22.2, 20.7, 17.9; MS
(1S,3R,4R)-2-[(S)-1-Phenylethylamino]-2-azabicyclo[2.2.1]heptane-3-car-
boxaldehyde (12): To a solution of oxalyl chloride (4.82 g, 38 mmol) in dry
CH2Cl2 (100 mL) at 788C was added a solution of DMSO (6.39 g,
82.6 mmol) in CH2Cl2 (10 mL) over 5 min, and the reaction mixture was
stirred for 10 min at 788C. The alcohol 11 (8.01 g, 34.4 mmol) was added
as a solution in CH2Cl2 (10 mL) over 5 min. The reaction mixture was
stirred for 15 min, and an excess of Et3N (17.4 g, 120 mmol) was added over
5 min. The cooling bath was removed and the temperature raised to room
temperature. Water (120 mL) was added and the phases were separated.
The aqueous phase was extracted with CH2Cl2 (3 Â 60 mL). The combined
organic extracts were washed with brine (60 mL) and dried over MgSO4.
Filtration and evaporation afforded a residue that was purified by flash
chromatography with Et2O/pentane (1/4) as eluent to give a yellow oil.
‡
‡
‡
(70 eV, EI): m/z (%): 274 (16) [M ], 105 (100) [C8H9 ], 104 (15) [C8H8 ],
103 (23) [C8H7 ]; C18H27NO (273.4): calcd: C 79.07, H 9.95, N 5.12; found: C
‡
78.91, H 10.05, N 5.15.
(1S,3R,4R)-2-[(S)-1-Phenylethylamino]-2-azabicyclo[2.2.1]hept-5-ene-3-
phenylketone (16a): Phenyl glyoxylate (2.0 g, 15 mmol) was dissolved in
CH2Cl2 (15 mL), containing 3 g of activated 4 Š molecular sieves. The
solution was cooled to 08C, and (S)-1-phenylethylamine (2.0 mL,
15.7 mmol) was added. After 30 min, the solution was cooled to 788C,
and trifluoroacetic acid (1.2 mL, 15.7 mmol), BF3 ´ Et2O (1.97 mL,
15.7 mmol), and cyclopentadiene (1.3 mL, 15.7 mmol) were added. The
reaction was kept at 788C for 5 h then allowed to warm to room
temperature. The molecular sieves were removed, and the reaction mixture
was washed with saturated aqueous NaHCO3 (10 mL), brine (10 mL), and
dried over Na2SO4. Concentration and purification by flash chromatog-
raphy (deactivated silica gel, pentane/Et2O 95/5 ± 80/20) afforded pure 16a.
Yield: 92%; Rf ˆ 0.38 (Et2O/pentane 1/4); [a]2D5 ˆ ‡78.6 (c ˆ 2.00 in
1
CHCl3); IR (neat): nÄ ˆ 1721, 2970, 3062 cm
;
1H NMR: d ˆ 9.00 (d,
3J(H,H) ˆ 2.90 Hz, 1H), 7.33 ± 7.17 (m, 5H), 3.81 (s, 1H), 3.52 (q, 3J(H,H) ˆ
6.56 Hz, 1H), 2.41 (brs, 2H), 1.99 ± 2.05 (m, 1H), 1.62 ± 1.73 (m, 2H), 1.44 ±
3
3
1.51 (m, 1H), 1.39 (d, J(H,H) ˆ 6.41 Hz, 3H), 1.32 (d, J(H,H) ˆ 9.77 Hz,
1H); 13C NMR: d ˆ 205.0, 144.8, 128.5, 127.9, 127.6, 75.7, 60.8, 58.3, 42.3,
‡
Yield: 42%; Rf ˆ 0.45 (Et2O/pentane 1/4); [a]2D5 ˆ ‡10.1 (c ˆ 1.00 in
36.4, 29.2, 22.6, 22.5; MS (70 eV, EI): m/z (%): 229 (10) [M ], 200 (40)
‡
‡
1
[C14H18N ], 105 (100) [C8H9 ]; C15H19NO (229.3): calcd: C 78.56, H 8.35, N
6.11, found: C 78.29, H 8.31, N 6.25.
CH2Cl2); IR (CHCl3): nÄ ˆ 3680, 3601, 2784, 1711, 1691, 1606 cm
;
1H NMR: d ˆ 7.45 ± 6.9 (m, 10H), 6.5 (s, 1H), 6.00 (dd, 3J(H,H) ˆ 6.3 Hz,
3J(H,H) ˆ 2.0 Hz, 1H), 3.20 (s, 1H), 4.40 (s, 1H), 3.15 (q, 3J(H,H) ˆ 6.6 Hz,
1H), 2.85 (s, 1H), 2.10 (d, 3J(H,H) ˆ 8.4 Hz, 1H), 1.45 (d, 3J(H,H) ˆ 6.9 Hz,
3H), 1.40 (d, 3J(H,H) ˆ 8.8 Hz, 1H); 13C NMR: d ˆ 145.1, 137.3, 136.1,
133.6, 132.2, 128.2, 128.1, 128.0, 127.5, 127.0, 121.3, 66.9, 64.2, 62.7, 49.3, 44.5,
22.7; MS (70 eV, EI): m/z (%): 304 (5) [M ], 105 (100) [C8H9 ], 104 (11)
[C8H8 ], 94 (17) [C6H7N ]; C21H21NO (303.4): calcd: C 83.13, H 6.98, N
4.62; found: C 83.21, H 6.87, N 4.78.
(1S,3R,4R)-2-[(S)-1-Phenylethylamino]-2-azabicyclo[2.2.1]heptane-3-(S)-
phenylmethanol (13a): At 788C phenylmagnesium bromide (0.45 mL,
1.33 mmol, 3m in Et2O) was added to a suspension of cerium chloride
(327 mg, 1.33 mmol) in dry THF (3 mL). After stirring for 1 h the mixture
was treated with a solution of 6 (100 mg, 0.44 mmol) in dry THF (2 mL),
and the reaction vessel was allowed to reach room temperature overnight.
Then brine (10 mL) was added and the product was extracted with CH2Cl2
(3 Â 10 mL). After drying over Na2SO4 the solvent was evaporated and the
residue was purified by flash chromatography. Yield: 98%; Rf ˆ 0.51
‡
‡
‡
‡
(1S,3R,4R)-2-[(S)-1-Phenylethylamino]-2-azabicyclo[2.2.1]hept-5-ene-3-
methylketone (16b): To a cooled solution (08C) of methyl glyoxylate
(17.82 g, 41 mmol, 40% solution in H2O) in H2O (20 mL) was added
dropwise an aqueous solution of (S)-1-phenylethylamine (10 g, 49.5 mmol)
in 2m HCl (41 mL). After stirring for 30 min cyclopentadiene (10.84 g,
164 mmol) was added, stirring was continued overnight, and the solution
was allowed to warm to room temperature. Then the reaction mixture was
adjusted to pH 8 with NaOH. Extraction with CH2Cl2 (3 Â 50 mL), drying
over Na2SO4, concentration under reduced pressure, and purification by
flash chromatography (deactivated silica gel, pentane/Et2O 95/5 ± 80/20)
afforded pure 16b. Yield: 31%; Rf ˆ 0.15 (Et2O/pentane 1/4); [a]2D5 ˆ ‡10.8
(c ˆ 0.50 in CHCl3); IR (CHCl3): nÄ ˆ 2875, 1721, 1693 cm 1; 1H NMR: d ˆ
7.40 ± 7.20 (m, 5H), 6.50 (dd, 3J(H,H) ˆ 5.5 Hz, 3J(H,H) ˆ 3.0 Hz, 1H), 6.30
(dd, 3J(H,H) ˆ 5.5 Hz, 3J(H,H) ˆ 1.5 Hz, 1H), 4.37 (s, 1H), 3.03 (q,
3J(H,H) ˆ 6.4 Hz, 1H), 2.85 (s, 1H), 2.35 (s, 1H), 2.00 (d, 3J(H,H) ˆ
8.5 Hz, 1H), 1.80 ± 1.20 (m, 1H), 1.64 (s, 3H), 1.45 (d, 3J(H,H) ˆ 6.7 Hz,
3H); 13C NMR: d ˆ 145.2, 136.5, 132.9, 128.3, 128.2, 127.3, 70.9, 63.8, 62.7,
(EtOAc/pentane 1/1); [a]D25
ˆ
83.4 (c ˆ 1.00 in CHCl3); IR (neat): nÄ ˆ
3421, 2972, 2871, 1494, 1451.6 cm 1; 1H NMR: d ˆ 7.45 ± 6.95 (m, 10H), 3.75
(s, 1H), 3.63 (q, 3J(H,H) ˆ 6.8 Hz, 1H), 3.10 (d, 3J(H,H) ˆ 6.4 Hz, 1H), 2.22
(d, 3J(H,H) ˆ 6.2 Hz, 1H), 1.90 ± 2.05 (m, 2H), 1.75 (brs, 1H), 1.20 ± 1.50
(m, 5H), 1.05 (d, 3J(H,H) ˆ 8.1 Hz, 2H); 13C NMR: d ˆ 186.8, 128.6, 127.8,
127.7, 126.3, 125.4, 107.9, 73.1, 71.9, 60.8, 58.7, 37.6, 36.1, 29.8, 22.5, 22.3; MS
‡
‡
‡
(70 eV, EI): m/z (%): 308 (<1) [M ], 106 (10) [C7H6O ], 105 (100) [C8H9 ];
C21H25NO (307.4): calcd: C 82.04, H 8.2, N 4.56; found: C 81.84, H 8.08, N
4.41.
(1S,3R,4R)-2-[(S)-1-Phenylethylamino]-2-azabicyclo[2.2.1]heptane-3-(S)-
methylmethanol (13b): At 788C methylmagnesium chloride (0.48 mL,
1.46 mmol, 3m in THF) was added to a suspension of cerium chloride
(360 mg, 1.46 mmol) in dry THF (3 mL). After stirring for 1 h the mixture
was treated with a solution of 6 (110 mg, 0.48 mmol) in dry THF (2 mL) and
the reaction vessel was allowed to reach room temperature overnight. Then
brine (10 mL) was added and the product was extracted with CH2Cl2 (3 Â
10 mL). After drying over Na2SO4 the solvent was evaporated and the
residue was purified by flash chromatography. Yield: 98%; Rf ˆ 0.40
‡
48.5, 44.7, 28.0, 22.0; MS (70 eV, EI): m/z (%): 242 (15) [M ], 198 (100)
‡
‡
[C14H16N ], 106 (96) [C8H9 ]; C16H19NO (241.3): calcd: C 79.63, H 7.94, N
5.80; found: C 79.78, H 8.07, N 6.05.
(EtOAc/pentane 1/1); [a]D25
ˆ
48.6 (c ˆ 1.00 in CHCl3); IR (CHCl3): nÄ ˆ
3695, 3383, 2962, 2874, 1602, 1309, 1277, 1054 cm 1; 1H NMR: d ˆ 7.40 ± 7.20
General procedure for alkene reduction: 1 g of the 2-azanorbornene 16a or
16b was dissolved in MeOH (25 mL), and the solution added to a stirred
suspension of Pd/C (100 mg, 10 wt%) and 200 mg K2CO3. After 3 h the
solvent was evaporated. Then the residue was dissolved in H2O (20 mL)
and extracted with CH2Cl2 (3 Â 20 mL). The combined organic extracts
were dried over Na2SO4 and evaporated to give the white, low-melting
solids 17a and 17b.
3
(m, 5H), 3.70 (s, 1H), 3.55 (q, J(H,H) ˆ 7.2 Hz, 1H), 3.20 (brs, 1H), 2.35
(d, 3J(H,H) ˆ 4.8 Hz, 1H), 2.20 (d, 3J(H,H) ˆ 4.8 Hz, 1H), 2.10 ± 1.80 (m,
2H), 1.70 ± 1.15 (m, 5H), 1.38 (d, 3J(H,H) ˆ 6.5 Hz, 3H), 0.90 (d, 3J(H,H) ˆ
6.9 Hz, 3H); 13C NMR: d ˆ 146.0, 128.5, 127.5, 127.5, 107.4, 72.9, 65.5, 60.9,
58.4, 36.6, 36.3, 29.8, 22.8, 22.4, 17.8; MS (70 eV, EI): m/z (%): 245 (23)
‡
‡
‡
[M ], 105 (38) [C8H9 ], 91 (100), 69 (30) [C4H5O ]; C16H23NO (245.4):
calcd: C 78.32, H 9.45, N 5.71; found: C, 78.21, H 9.62, N 5.84.
(1S,3R,4R)-2-[(S)-1-Phenylethylamino]-2-azabicyclo(2.2.1(heptane-3-
(1S,3R,4R)-2-[(S)-1-Phenylethylamino]-2-azabicyclo[2.2.1]heptane-3-(S)-
isopropylmethanol (13c): At 788C the Grignard reagent derived from
isopropyl bromide (0.27 mL, 3 mmol) and magnesium (80 mg, 3 mmol) in
dry THF (5 mL) was added to a suspension of cerium chloride (750 mg,
3 mmol) in dry THF (10 mL). After stirring for 1 h the mixture was treated
with a solution of 6 (229 mg, 1 mmol) in dry THF (2 mL) and the reaction
vessel was allowed to reach room temperature overnight. Then brine
phenylketone (17a): Yield: 95%; Rf ˆ 0.2 (Et2O/pentane 1/4); [a]D25
ˆ
‡0.6 (c ˆ 0.50 in CHCl3); IR (CHCl3): nÄ ˆ 3695, 2963, 2874, 1691, 1600,
1
1312, 1228 cm 1; H NMR: d ˆ 7.40 ± 6.90 (m, 10H), 3.90 (s, 1H), 3.60 (q,
3
3J(H,H) ˆ 6.5 Hz, 1H), 3.51 (s, 1H), 2.25 (d, J(H,H) ˆ 5.6 Hz, 1H), 2.10 ±
3
2.21 (m, 2H), 1.42 ± 1.80 (m, 3H), 1.40 (d, J(H,H) ˆ 6.7 Hz, 3H), 1.31 (d,
3J(H,H) ˆ 6.7 Hz, 1H); 13C NMR: d ˆ 200.6, 144.8, 137.0, 132.0, 128.3,
128.0, 128.0, 127.4, 127.1, 72.3, 61.4, 58.2, 43.2, 35.3, 29.6, 22.6, 22.1; MS
Chem. Eur. J. 1999, 5, No. 6
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0506-1697 $ 17.50+.50/0
1697

P. G. Andersson et al.
FULL PAPER
‡
‡
3
(70 eV, EI): m/z (%): 306 (16) [M ], 105 (100) [C8H9 ]; C21H23NO (305.4):
3J(H,H) ˆ 5.7 Hz, 1H), 2.75 (d, J(H,H) ˆ 5.7 Hz, 1H), 2.46 (s, 1H), 1.85 ±
1.10 (m, 7H), 0.90 (d, 3J(H,H) ˆ 6.8 Hz, 3H), 0.86 (d, 3J(H,H) ˆ 6.8 Hz,
3H); 13C NMR: d ˆ 77.1, 63.0, 55.7, 36.9, 35.0, 31.2, 29.5, 29.2, 19.4, 17.4; MS
calcd: C 82.59, H 7.59, N 4.59; found: C 82.35, H 7.44, N 4.72.
(1S,3R,4R)-2-[(S)-1-Phenylethylamino]-2-azabicyclo[2.2.1]heptane-3-me-
‡
‡
thylketone (17b): Yield: 95%; Rf ˆ 0.5 (Et2O/pentane 1/4); [a]D25
ˆ
23.2
(70 eV, EI): m/z (%): 168 (9) [M
C10H19NO (169.3): calcd: C 70.96, H 11.31, N 8.27; found: C 71.10, H 11.45,
N 8.21.
H], 96 (28) [C6H10N ], 68 (100);
(c ˆ 0.50 in CHCl3); IR (CHCl3): nÄ ˆ 3695, 2964, 2874, 1712, 1693, 1602,
1
1314, 1191 cm
;
1H NMR: d ˆ 7.40 ± 7.10 (m, 5H), 3.80 (s, 1H), 3.45 (q,
3J(H,H) ˆ 6.6 Hz, 1H), 2.65 (s, 1H), 2.21 (d, J(H,H) ˆ 3.6 Hz, 1H), 2.11 ±
1.90 (m, 2H), 1.82 ± 1.21 (m, 4H), 1.50 (s, 3H), 1.36 (d, 3J(H,H) ˆ 6.5 Hz,
3H); 13C NMR: d ˆ 210.3, 186.8, 145.0, 128.4, 128.2, 127.4, 76.6, 61.0, 58.1,
(1S,3R,4R)-2-Azabicyclo[2.2.1]heptane-3-(R)-phenylmethanol
(19a):
3
Yield: 96%; [a]2D5 ˆ ‡30.4 (c ˆ 1.00 in CHCl3); IR (CHCl3): nÄ ˆ 3680,
3601, 2874, 1730, 1691, 1607 cm 1; 1H NMR: d ˆ 7.40 ± 7.20 (m, 5H), 4.11 (d,
3J(H,H) ˆ 8.5 Hz, 1H), 3.55 (s, 1H), 3.32 (brs, 2H), 2.78 (d, 3J(H,H) ˆ
8.2 Hz, 1H), 2.01 (s, 1H), 1.80 ± 1.22 (m, 5H), 1.20 (d, 3J(H,H) ˆ 10.2 Hz,
1H); 13C NMR: d ˆ 143.5, 128.4, 127.5, 126.7, 107.3, 75.3, 67.7, 56.2, 39.0,
‡
42.3, 35.5, 29.9, 27.3, 22.1, 22.0; MS (70 eV, EI): m/z (%): 244 (8) [M ], 105
‡
‡
(100) [C8H9 ], 96 (16) [C6H9N ]; C16H21NO (243.3): calcd: C 78.97, H 8.7, N
5.74; found: C 79.10, H 8.85, N 5.58.
‡
‡
34.4, 28.3; MS (70 eV, EI): m/z (%): 204 (100) [M ], 186 (19) [C13H15N ], 96
General procedure for ketone reduction: To a suspension of LiAlH4
(250 mg, 6.55 mmol) in THF (20 mL) at 788C was added dropwise a
solution (10 mL) of the ketone (1 g, 3.28 mmol) in THF by syringe. After
30 min the cooling bath was removed, and stirring was continued for 1 h.
The reaction was quenched by adding H2O (0.25 mL) and 2m NaOH
(0.5 mL) and then filtered through Celite. Evaporation afforded a residue,
which was purified by flash chromatography on deactivated silica gel to
give the protected b-amino alcohol.
‡
(74) [C6H10N ]; C13H17NO (203.3): calcd: C 76.81, H 8.43, N 6.89; found: C
76.73, H 8.56, N 6.93.
General procedure for carbamate formation: At 08C phosgene (0.46 mL,
1.93m solution in toluene, 0.89 mmol) was added to a solution of the b-
amino alcohol (105 mg, 0.75 mmol) in THF (20 mL). After 10 min Et3N
(3.0 equiv) was added. Then after 10 min the cooling bath was removed,
and stirring was continued at room temperature for 30 min. Evaporation of
the solvent under reduced pressure afforded the crude carbamates, which
were purified by flash chromatography on silica gel.
(1S,3R,4R)-2-[(S)-1-Phenylethylamino]-2-azabicyclo[2.2.1]heptane-3-
(R)-phenylmethanol (18a): Yield: 71%; Rf ˆ 0.51 (EtOAc/pentane 1/1);
[a]2D5
ˆ
5.28 (c ˆ 2.40 in CHCl3); IR (neat): nÄ ˆ 3216, 2958, 2875, 1603,
(1S,3R,4R)-2-Azabicyclo[2.2.1]heptane-3-(S)-phenylmethanol N,O-carba-
1442, 1309, 1204, 1061 cm 1; 1H NMR: d ˆ 7.21 ± 7.40 (m, 5H), 7.05 ± 7.15 (m,
3H), 6.48 ± 6.59 (m, 2H), 3.75 (d, 3J(H,H) ˆ 5.8 Hz, 1H), 3.68 (s, 1H), 3.52
(q, 3J(H,H) ˆ 6.1 Hz, 1H), 2.30 (d, 3J(H,H) ˆ 5.1 Hz, 1H), 2.21 ± 2.05 (brm,
1H), 2.26 (s, 1H), 1.60 ± 1.80 (m, 2H), 1.45 (d, 3J(H,H) ˆ 6.80 Hz, 3H),
1.35 ± 1.20 (m, 3H); 13C NMR: d ˆ 186.8, 128.6, 127.8, 127.7, 126.3, 125.4,
107.9, 73.1, 71.9, 60.8, 58.7, 37.6, 36.1, 29.8, 22.5, 22.3; MS (70 eV, EI): m/z
mate (15a): Yield: 60%; Rf ˆ 0.70 (pentane/EtOAc 1/1); [a]2D5 ˆ ‡64.0 (c ˆ
1
1.00 in CHCl3); IR (CHCl3): nÄ ˆ 3659, 2954, 1730, 1755, 1602, 1319 cm
;
1H NMR: d ˆ 7.52 ± 7.30 (m, 5H), 4.95 (d, 3J(H,H) ˆ 8.0 Hz, 1H), 4.35 (s,
1H), 3.45 (d, 3J(H,H) ˆ 8.0 Hz, 1H), 2.60 (s, 1H), 1.80 ± 1.35 (m, 8H);
13C NMR: d ˆ 138.4, 128.9, 128.8, 125.8, 83.5, 70.2, 61.0, 40.7, 36.4, 27.8, 27.6;
‡
‡
MS (70 eV, EI): m/z (%): 229 (52) [M ], 184 (77) [C13H15N ], 168 (100), 95
‡
‡
‡
‡
(%): 307 (2) [M ], 274 (100) [C20H20N ], 179 (75), 105 (18) [C8H9 ], 95 (23)
[C6H9N ]; C21H25NO (307.4): calcd: C 82.04, H 8.2, N 4.56; found: C 82.17,
(48) [C6H9N ]; C14H15NO2 (229.3): calcd: C 73.34, H 6.59, N 6.11; found: C
‡
73.21, H 6.80, N 6.35.
H 8.13, N 4.65.
(1S,3R,4R)-2-Azabicyclo[2.2.1]heptane-3-(S)-methylmethanol N,O-carba-
(1S,3R,4R)-2-[(S)-1-Phenylethylamino]-2-azabicyclo[2.2.1]heptane-3-
(R)-methylmethanol (18b): Yield: 61%; Rf ˆ 0.40 (EtOAc/pentane 1/1);
mate (15b): Yield: 30%; Rf ˆ 0.70 (pentane/EtOAc 1/1); [a]2D5 ˆ ‡12.7
1
(c ˆ 1.00 in CHCl3); IR (CHCl3): nÄ ˆ 3054, 2988, 2307, 1424, 1271 cm
;
[a]2D5
1080 cm
ˆ
50.4 (c ˆ 1.00 in CH2Cl2); IR (CHCl3): nÄ ˆ 3260, 3601, 1606,
1H NMR: d ˆ 4.85 (dq, 3J(H,H) ˆ 9.2 Hz, 3J(H,H) ˆ 9.2 Hz, 1H), 4.26 (s,
1H), 3.53 (d, 3J(H,H) ˆ 9.1 Hz, 1H), 2.52 (s, 1H), 1.75 ± 1.55 (m, 3H), 1.20 ±
1
;
1H NMR: d ˆ 7.40 ± 7.20 (m, 5H) , 3.64 (s, 1H), 3.52 (q,
3J(H,H) ˆ 6.54 Hz, 1H), 2.83 (dq, 3J(H,H) ˆ 4.14 Hz, 3J(H,H) ˆ 2.2 Hz,
1H), 2.11 (d, 3J(H,H) ˆ 4.08 Hz, 2H), 1.99 (d, 3J(H,H) ˆ 3.84 Hz, 1H),
1.60 ± 1.82 (m, 2H), 1.45 (d, 3J(H,H) ˆ 6.54 Hz, 3H), 1.45 ± 1.20 (m, 3H),
0.33 (d, 3J(H,H) ˆ 6.36 Hz, 3H); 13C NMR: d ˆ 144.9, 128.7, 128.4, 128.3,
127.6, 127.5, 72.1, 68.0, 60.9, 58.7, 42.9, 35.3, 27.8, 23.4, 22.3, 21.0; MS (70 eV,
1.40 (m, 3H), 1.21 (d, J(H,H) ˆ6.7 Hz, 1H); 13C NMR: d ˆ 163.2, 106.5,
3
77.2, 74.9, 64.9, 60.6, 38.1, 37.6, 28.7, 27.2, 16.3; MS (70 eV, EI): m/z (%): 168
‡
(63) [M ], 105 (100); HRMS calcd for C9H13NO2: 167.0946; found:
167.0945.
(1S,3R,4R)-2-Azabicyclo[2.2.1]heptane-3-(R)-phenylmethanol N,O-car-
bamate (20a): Yield: 60%; Rf ˆ 0.70 (pentane/EtOAc 1/1); [a]2D5 ˆ ‡6.1
‡
EI): m/z (%): 246 (21) [M ], 79 (100); C16H23NO (245.4): calcd: C 78.32, H
9.45, N 5.71; found: C 78.53, H 9.50, N 5.58.
(c ˆ 1.00 in CHCl3); IR (CHCl3): nÄ ˆ 3659, 2953, 2881, 1756, 1728, 1603,
1
General procedure for debenzylation: To a solution of the protected b-
amino alcohol (1 g) in EtOH (50 mL) was added Pd(OH)2/C (200 mg,
20 wt%). The mixture was hydrogenated for 48 h at 20 atm pressure. The
completeness of the reaction was monitored by NMR spectroscopy. The
catalyst was filtered off on a bed of Celite, washed with CH2Cl2 (20 mL),
and the combined filtrates were evaporated to dryness. The residue was
triturated with pentane and dried to give the deprotected b-amino alcohols
as off-white powders.
1316, 1236 cm
;
1H NMR: d ˆ 7.50 ± 7.20 (m, 5H), 5.75 (d, 3J(H,H) ˆ
9.2 Hz, 1H), 4.25 (s, 1H), 3.85 (d, 3J(H,H) ˆ 9.2 Hz, 1H), 2.05 (d,
3J(H,H) ˆ 3.2 Hz, 1H), 1.70 ± 1.35 (m, 5H), 1.00 (s, 1H); 13C NMR: d ˆ
163.1, 136.3, 128.4, 128.1, 125.2, 79.2, 66.3, 59.8, 38.8, 36.7, 28.4, 27.7; MS
(70 eV, EI): m/z (%): 229 (55) [M ], 184 (66) [C13H15N ], 168 (100), 95 (40)
[C6H9N ]; C14H15NO2 (229.3): calcd: C 73.34, H 6.59, N 6.11; found: C
‡
‡
‡
73.42, H 6.55, N 5.97.
(1S,3R,4R)-2-Azabicyclo[2.2.1]heptane-3-(R)-methylmethanol N,O-car-
(1S,3R,4R)-2-Azabicyclo[2.2.1]heptane-3-(S)-phenylmethanol
(14a):
bamate (20b): Yield: 73%; Rf ˆ 0.70 (pentane/EtOAc 1/1); [a]2D5 ˆ ‡1.1
Yield: 96%; [a]2D5 ˆ ‡13.2 (c ˆ 0.50 in CHCl3); IR (neat): nÄ ˆ 3312,
(c ˆ 0.28 in CHCl3); IR (CHCl3): nÄ ˆ 3055, 2987, 2306, 1424, 1271 cm
;
1
1
2933 cm
;
1H NMR: d ˆ 7.38 ± 7.10 (m, 5H), 4.27 (d, 3J(H,H) ˆ 6.2 Hz,
1H NMR: d ˆ 4.23 (s, 1H), 4.11 (dq, 3J(H,H) ˆ 8.0, 3J(H,H) ˆ 8.0 Hz, 1H),
2.07 (s, 1H), 2.44 (s, 1H), 1.70 ± 1.30 (m, 5H), 1.50 (dd, 3J(H,H) ˆ 8.0,
3J(H,H) ˆ 8.0 Hz, 3H); 13C NMR: d ˆ 78.6, 69.5, 60.5, 40.4, 36.4, 27.7, 27.7,
1H), 3.40 (s, 1H), 2.85 (d, 3J(H,H) ˆ 6.2 Hz, 1H), 2.25 (brs, 3H), 1.65 ± 1.10
(m, 5H), 1.05 (d, 3J(H,H) ˆ 8.1 Hz, 1H); 13C NMR: d ˆ 128.3, 127.3, 126.6,
107.4, 107.3, 75.6, 65.9, 55.8, 37.6, 34.8, 32.7, 29.4; MS (70 eV, EI): m/z (%):
‡
20.3; MS (70 eV, EI): m/z (%): 168 (62) [M ], 105 (100); C9H13NO2 (167.2):
‡
‡
202 (90) [M ], 186 (18) [C13H17N ], 69 (100); C13H17NO (203.3): calcd: C
calcd: C 64.65, H 7.84, N 8.28; found: C 64.82, H 7.71, N 8.24.
76.81, H 8.43, N 6.89; found: C 76.88, H 8.52, N 6.85.
General procedure for the benzylation of the NH b-amino alcohols:
Benzylations of NH b-amino alcohols were performed according to a
literature procedure.[11]
(1S,3R,4R)-2-Azabicyclo[2.2.1]heptane-3-(S)-methylmethanol
(14b):
Yield: 93%; [a]2D5 ˆ ‡17.1 (c ˆ 0.50 in CHCl3); IR (neat): nÄ ˆ 3268, 2954,
1
2364 cm
;
1H NMR: d ˆ 4.63 (brs, 2H), 3.50 (q, 3J(H,H) ˆ 6.4 Hz, 1H),
(1S,3R,4R)-2-(Benzylamino)-2-azabicyclo[2.2.1]heptane-3-(S)-phenylme-
3.49 (s, 1H), 2.70 (s, 1H), 2.59 (d, 3J(H,H) ˆ 7.1 Hz, 1H), 2.52 (s, 1H),
1.78 ± 1.20 (m, 5H), 1.17 (d, 3J(H,H) ˆ 7.9 Hz, 3H), 1.09 (d, 3J(H,H) ˆ
1.1 Hz, 1H); 13C NMR: d ˆ 69.4, 66.7, 55.8, 36.9, 34.8, 32.3, 29.3, 19.4; MS
thanol (5): Yield: 61%; Rf ˆ 0.30 (pentane/EtOAc 4/1); [a]D25
ˆ
6.1 (c ˆ
0.4 in CH2Cl2); IR (CH2Cl2): nÄ ˆ 3602, 3408, 2873, 1606, 1495, 1070,
1
1014 cm
;
1H NMR: d ˆ 7.38 ± 7.21 (m, 10H), 4.39 (d, 3J(H,H) ˆ 4.8 Hz,
‡
‡
1H), 3.66 (brs, 2H), 3.30 ± 3.50 (m, 1H), 3.26 (brs, 1H), 2.30 (d, 3J(H,H) ˆ
4.8 Hz, 1H), 2.05 ± 1.90 (m, 1H), 1.88 (brd, 3J(H,H) ˆ 9.2 Hz, 1H), 1.52 ±
1.38 (m, 1H), 1.37 ± 1.21 (m, 1H), 1.20 ± 1.08 (m, 1H), 1.08 (brd, 3J(H,H) ˆ
9.6 Hz, 1H), 1.05 (brs, 1H); 13C NMR: d ˆ 141.7, 139.4, 128.9, 128.5, 128.3,
128.0, 127.1, 126.8, 125.9, 73.5, 72.7, 58.9, 53.9, 38.2, 36.4, 30.0, 21.9; MS
(70 eV, EI): m/z (%):141 (18) [M ], 96 (47) [C6H10N ], 68 (100); C8H15NO
(141.2): calcd: C 68.04, H 10.71, N 9.92; found: C 67.89, H 10.77, N 9.95.
(1S,3R,4R)-2-Azabicyclo[2.2.1]heptane-3-(S)-isopropylmethanol
(14c):
Yield: 95%; [a]2D5
ˆ
10.0 (c ˆ1.00 in CHCl3); IR (CHCl3): nÄ ˆ 3680,
1
3601, 2873, 1730, 1607 cm 1; H NMR: d ˆ 3.51 (brs, 3H), 3.02 (pseudo-t,
1698
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0506-1698 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 6

Asymmetric Addition
1692±1699
‡
‡
(70 eV, EI): m/z (%): 293 (6) [M ], 186 (17) [C13H14N ], 158 (100); HRMS
[4] a) K. Soai, A. Ookawa, T. Kaba, K. Ogawa, J. Am. Chem. Soc. 1987,
109, 7111; b) E. J. Corey, F. J. Hannon, Tetrahedron Lett. 1987, 28,
5233; c) E. J. Corey, P. Yuen, F. J. Hannon, D. A. Wierda, J. Org.
Chem. 1990, 55, 784; d) M. Shi, Y. Satoh, T. Makihara, Y. Masaki,
Tetrahedron: Asymmetry 1995, 6, 2109.
[5] M. J. Södergren, P. G. Andersson, Tetrahedron Lett. 1996, 37, 7577.
[6] D. A. Alonso, D. Guijarro, P. Pinho, O. Temme, P. G. Andersson, J.
Org. Chem. 1998, 63, 2749.
[7] P. Pinho, D. Guijarro, P. G. Andersson, Tetrahedron 1998, 54, 7897.
[8] a) K. Soai, T. Hatanaka, J. Chem. Soc. Chem. Commun. 1992, 1097;
b) W. B. Jennings, C. J. Lovely, Tetrahedron 1991, 47, 5561.
[9] D. Guijarro, P. Pinho, P. G. Andersson, J. Org. Chem. 1998, 63, 2530.
[10] M. Yamakawa, R. Noyori, J. Am. Chem. Soc. 1995, 117, 6327 ± 6335.
[11] a) P. G. Andersson, D. Guijarro, D. Tanner, J. Org. Chem. 1997, 62,
7364; b) P. G. Andersson, D. Guijarro, D. Tanner, Synlett 1996, 727.
[12] B3PW91 calculations on the aziridinoalcohol-catalyzed analogue of
the reaction[11a] showed that the same mechanism is in operation also
in this case (see Supporting Information).
calcd for C20H21NO: 293.1779; found: 293.16780.
(1S,3R,4R)-2-(Benzylamino)-2-azabicyclo[2.2.1]heptane-3-(S)-methylme-
thanol (6): Yield: 65%; Rf ˆ 0.43 (pentane/EtOAc 7/3); [a]2D4 ˆ ‡38.6 (c ˆ
1.1 in CH2Cl2); IR (CH2Cl2): nÄ ˆ 3623, 3408, 2872, 2807, 1754, 1726, 1704,
1691, 1679, 1051 cm 1; 1H NMR: d ˆ 7.40 ± 7.20 (m, 5H), 3.75 (d, 3J(H,H) ˆ
13.2 Hz, 1H), 3.65 (d, 3J(H,H) ˆ 13.2 Hz, 1H), 3.51 ± 3.45 (m, 1H), 3.19
(brs, 1H), 2.95 ± 3.20 (m, 1H), 2.41 (brs, 1H), 2.10 ± 1.85 (m, 2H), 1.82 ± 1.68
(m, 1H), 1.62 ± 1.56 (m, 1H), 1.40 ± 1.10 (m, 3H), 1.14 (d, 3J(H,H) ˆ 6.8 Hz,
3H); 13C NMR: d ˆ 128.7, 128.3, 127.0, 73.3, 66.4, 58.4, 54.0, 37.3, 36.7, 30.3,
‡
‡
22.0, 18.4; MS (70 eV, EI): m/z (%): 232 (5) [M ], 186 (82) [C11H12N ], 158
(100); HRMS calcd for C15H21NO: 231.1623; found: 231.1623.
(1S,3R,4R)-2-(Benzylamino)-2-azabicyclo[2.2.1]heptane-3-(S)-isopropyl-
methanol (7): Yield: 64%; Rf ˆ 0.46 (pentane/EtOAc 7/3); [a]2D4 ˆ ‡38.5
(c ˆ 1.2 in CH2Cl2); IR (CH2Cl2): nÄ ˆ 3623, 3407, 2872, 2808, 1721, 1709,
1
1027 cm
;
1H NMR: d ˆ 7.40 ± 7.20 (m, 5H), 3.73 (d, 3J(H,H) ˆ 13.2 Hz,
3
1H), 3.61 (d, J(H,H) ˆ 13.2 Hz, 1H), 3.18 (brs, 1H), 3.00 ± 2.90 (m, 1H),
2.38 (brs, 1H), 2.11 (brs, 1H), 2.02 ± 1.92 (m, 1H), 1.80 ± 1.54 (m, 3H),
1.38 ± 1.20 (m, 2H), 1.10 (d, 3J(H,H) ˆ 9.6 Hz, 1H), 1.05 (d, 3J(H,H) ˆ
6.8 Hz, 3H), 0.82 (d, 3J(H,H) ˆ 6.8 Hz, 3H); 13C NMR: d ˆ 128.9, 128.3,
127.0, 75.3, 70.0, 57.9, 53.3, 37.4, 36.9, 30.9, 30.6, 21.7, 20.7, 18.3; MS (70 eV,
[13] a) L. Stella, H. Abraham, J. Feneu-Dupont, B. Tinant, J. P. Declercq,
Tetrahedron Lett. 1990, 31, 2603; b) H. Abraham, L. Stella, Tetrahe-
dron 1992, 48, 9707.
[14] M. J. Södergren, P. G. Andersson, J. Am. Chem. Soc. 1998, 120, 10760.
[15] a) M. T. Reetz, Angew. Chem. 1991, 103, 1559; Angew. Chem. Int. Ed.
Engl. 1991, 30, 1531; b) M. T. Reetz, M. W. Drewes, A. Schmitz,
Angew. Chem. 1987, 99, 1186; Angew. Chem. Int. Ed. Engl. 1987, 26,
1141; c) W. C. Still, J. A. Schneider, Tetrahedron Lett. 1980, 21, 1035.
[16] L. A. Paquette in Encyclopedia of Reagents for Organic Synthesis,
Vol. 2 (Ed.: L. A. Paquette), Wiley, New York, 1995, p. 1031.
[17] T. Imamoto, N. Takiyama, K. Nakamura, T. Hatajima, Y. Kamiya, J.
Am. Chem. Soc. 1989, 111, 4392.
‡
‡
EI): m/z (%): 259 (<1) [M ], 186 (100) [C13H16N ]; HRMS calcd for
C17H25NO: 259.1936; found: 259.1935.
(1S,3R,4R)-2-(Benzylamino)-2-azabicyclo[2.2.1]heptane-3-(R)-phenylme-
thanol (8): Yield: 61%; Rf ˆ 0.42 (pentane/EtOAc 4/1); [a]D25
ˆ
67.2 (c ˆ
0.64 in CH2Cl2); IR (CH2Cl2): nÄ ˆ 3603, 3408, 2873, 1606, 1495, 1070,
1014 cm 1; H NMR: d ˆ 7.37 ± 7.23 (m, 10H), 4.35 ± 4.80 (m, 1H), 4.30 (d,
1
3J(H,H) ˆ 6.8 Hz, 1H), 3.77 (d, 3J(H,H) ˆ 13.3 Hz, 1H), 3.60 (d, 3J(H,H) ˆ
13.3 Hz, 1H), 3.23 (brs, 1H), 2.25 ± 2.34 (m, 2H), 2.11 ± 1.97 (m, 1H), 1.85
(brd, 3J(H,H) ˆ 9.7 Hz, 1H), 1.70 ± 1.55 (m, 1H), 1.35 ± 1.20 (m, 3H);
13C NMR: d ˆ 144.6, 139.0, 128.7, 128.4, 128.2, 127.0, 125.9, 75.4, 75.4, 74.8,
[18] Ligands 13a ± c and 18a ± b did not catalyze the addition reaction, and
no signals for the product could be observed in the NMR spectra of
the crude reaction mixtures.
‡
57.9, 55.4, 42.1, 35.6, 28.4, 22.3; MS (70 eV, EI): m/z (%): 294 (3) [M ], 186
‡
‡
(7) [C13H14N ], 91 (100) [C7H7 ]; HRMS calcd for C20H21NO: 293.1779;
found: 293.1678.
[19] O. Mitsunobu, Synthesis 1981, 1.
[20] a) H. Nakano, N. Kumagai, C. Kabuto, H. Matsuzaki, H. Hongo,
Tetrahedron: Asymmetry 1995, 6, 1233; b) H. Nakano, N. Kumagai, C.
Kabuto, H. Matsuzaki, H. Hongo, Tetrahedron: Asymmetry 1997, 8,
1391.
[21] a) A. D. Becke, J. Chem. Phys. 1993, 98 5648 ± 5652; b) J. P. Perdew, Y.
Wang, Phys. Rev. B 1992, 45, 13244.
[22] Gaussian 94, Revision D.1; M. J. Frisch, G. W. Trucks, H. B. Schlegel,
P. M. W. Gill, B. G. Johnson, M. A. Robb, J. R. Cheeseman, T. Keith,
G. A. Petersson, J. A. Montgomery, K. Raghavachari, M. A. Al-
Laham, V. G. Zakrzewski, J. V. Ortiz, J. B. Foresman, J. Cioslowski,
B. B. Stefanov, A. Nanayakkara, M. Challacombe, C. Y. Peng, P. Y.
Ayala, W. Chen, M. W. Wong, J. L. Andres, E. S. Replogle, R.
Gomperts, R. L. Martin, D. J. Fox, J. S. Binkley, D. J. Defrees, J.
Baker, J. P. Stewart, M. Head-Gordon, C. Gonzalez, J. A. Pople,
Gaussian, Inc., Pittsburgh PA, 1995.
Acknowledgments
We are grateful to Mikael Södergren for the NOE data. We thank the
Swedish Natural Science Research Council, the Foundation for Strategic
Research (SSF), STINT, Astra-Arcus, and the Swedish Research Council
for Engineering Sciences (TFR) for financial support. Access to super-
computer time was provided by the Swedish Council for High Performance
Computing (HPDR) and Parallelldatorcentrum (PDC), Royal Institute of
Technology.
[23] A. V. Bedekar, E. B. Koroleva, P. G. Andersson, J. Org. Chem. 1997,
62, 2518.
[1] R. Noyori, M. Kitamura, Angew. Chem. 1991, 103, 34; Angew. Chem.
Int. Ed. Engl. 1991, 30, 49.
[2] K. Soai, S. Niwa, Chem. Rev. 1992, 92, 833.
[3] a) K. Soai, A. Oakawa, J. Chem. Soc. Chem. Commum. 1986, 412;
b) K. Soai, A. Oakawa, J. Chem. Soc. Perkin Trans. 1 1987, 1465.
Received: November 6, 1998 [F1429]
Chem. Eur. J. 1999, 5, No. 6
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0506-1699 $ 17.50+.50/0
1699