Chiral amine-imine ligands based on trans-2,5-disubstituted pyrrolidines and their application in the palladium-catalyzed allylic alkylation

Article abstract of DOI:10.1016/j.tetasy.2009.07.010

A series of amine-imine bidentate ligands based on a trans-2,5-disubstituted pyrrolidine and pyridine moieties have been prepared. The use of these ligands in the palladium-catalyzed allylic alkylation reaction of rac-(E)-1,3-diphenylprop-2-enyl acetate is reported. The results suggest that these ligands are good catalyst precursors for the reaction. Electronic modification on the pyridine ring of the ligands does not have a significant effect on the enantioselectivity of the reaction but does on the reaction rate, while structural modification on either the pyridine or the pyrrolidine moiety affords dramatic changes on the outcome of the stereochemistry. Evidence from various studies suggested that during the palladium-catalyzed allylic alkylation reaction, nucleophilic attack onto the 1,3-diphenylallyl moiety in the transition state occurs mainly trans to the pyridine ring of the less stable conformation of the palladium complexes.

Full text of DOI:10.1016/j.tetasy.2009.07.010

Tetrahedron: Asymmetry 20 (2009) 1672–1682
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Chiral amine–imine ligands based on trans-2,5-disubstituted pyrrolidines
and their application in the palladium-catalyzed allylic alkylation
Hongfeng Chen a, , James A. Sweet b, Kin-Chung Lam c, Arnold L. Rheingold c,à, Dominic V. McGrath a,
*
a Department of Chemistry, University of Arizona, Tucson, AZ 85721, USA
b Department of Chemistry, University of Connecticut, Storrs, CT 06269, USA
c Department of Chemistry, University of Delaware, Newark, DE 19716, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of amine–imine bidentate ligands based on a trans-2,5-disubstituted pyrrolidine and pyridine
moieties have been prepared. The use of these ligands in the palladium-catalyzed allylic alkylation
reaction of rac-(E)-1,3-diphenylprop-2-enyl acetate is reported. The results suggest that these ligands
are good catalyst precursors for the reaction. Electronic modification on the pyridine ring of the ligands
does not have a significant effect on the enantioselectivity of the reaction but does on the reaction rate,
while structural modification on either the pyridine or the pyrrolidine moiety affords dramatic changes
on the outcome of the stereochemistry. Evidence from various studies suggested that during the palla-
dium-catalyzed allylic alkylation reaction, nucleophilic attack onto the 1,3-diphenylallyl moiety in the
transition state occurs mainly trans to the pyridine ring of the less stable conformation of the palladium
complexes.
Received 6 May 2009
Accepted 2 July 2009
Available online 28 July 2009
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
from both the pyridine and pyrrolidine moieties are different elec-
tronically and are assumed to provide different binding properties
Since its introduction by Whitesell in 1977, C2-symmetric trans-
2,5-disubstituted pyrrolidines1 have played an important role as
useful chiral auxiliaries in asymmetric synthesis.1–13 Despite their
great success as chiral auxiliaries in asymmetric reactions, incorpo-
ration of trans-2,5-disubstituted pyrrolidines in chiral ligands for
transition metal-catalyzed asymmetric reactions is relatively
rare.14–20 We believe that the potential of this chiral building block
to influence asymmetric transformations has not been fully inves-
tigated. As part of an ongoing project on the design and synthesis
of ligands for asymmetric catalytic reactions, we developed a
modular synthetic strategy for preparing amine–imine ligands
based on trans-2,5-disubstituted pyrrolidines.16 Since our initial
report of this class of ligands,16 several examples of both
trans-2,5-disubstituted pyrrolidine-containing20–26 and pyridine–
amine27–30 and ligands have been reported, in addition to many
tertiary diamines.31
to transition metals. Although pyridine is a strong electron-donat-
ing ligand, the delocalized -system provides a tool for tuning the
p
electronic nature of this moiety with substituents. While the differ-
ent donor abilities of the pyridine and pyrrolidine rings can serve
as an electronic differentiator for transition metal-catalyzed asym-
metric reactions, the trans-2,5-disubstituted pyrrolidine moiety
can provide a chiral influence for asymmetric discrimination on
prochiral substrates.
R
R'
N
n
N
R
A: n = 1, 2
To study the scope and efficiency of this new type of bidentate
ligand, we modified the ligands in three ways: modification of the
pyrrolidine ring to provide different chiral environments for asym-
metric recognition; variation of the substituents on the pyridine
ring to create different electronic properties for the ligands; and
alteration of the linkage between the pyridine and the pyrrolidine
rings to provide different chelation geometry for the catalysts.
Herein we report in detail the synthesis and application of this ser-
ies of pyridine–pyrrolidine ligands in the palladium-catalyzed
allylic alkylation reaction of (rac)-(E)-1,3-diphenylprop-2-enyl ace-
tate with dimethyl malonate.
This type of ligand is constructed by linking a pyridine ring to
the pyrrolidine ring as chiral bidentate ligands A. Nitrogen atoms
* Corresponding author. Tel.: +1 520 626 4690; fax: +1 520 621 8407.
E-mail address: mcgrath@u.arizona.edu (D.V. McGrath).
Present address: Wilmington PharmaTech Company, One Innovation Way, Suite
302, Newark, DE 19711, USA.
à
To whom correspondence for the X-ray crystallographic results should be
addressed: Department of Chemistry and Biochemistry, University of California, San
Diego, La Jolla, CA 92093, USA.
0957-4166/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2009.07.010

H. Chen et al. / Tetrahedron: Asymmetry 20 (2009) 1672–1682
1673
O
OH
2. Results and discussion
2.1. Synthetic strategy
1. (-)-DIP-Cl, -78 o
C
R
R
R
R
2. HN(CH2CH2OH)2
O
OH
1: R = p-Tol
6: R = p-Tol, 83%
7: R = 4-t-BuPh, 85%
8: R = 2-naphthyl, 19%
9: R = 4-OMePh, decomposed
10: R = mesityl, no reaction
Retrosynthetic analysis reveals three different strategies for the
construction of this series of new chiral ligands (Scheme 1). In the
first route, a pyridine with a pendant amine is used to cyclize the
dimesylate of a chiral 1,4-diol to form the desired ligands (method
A). In the second route, a preformed trans-2,5-disubstituted pyrrol-
idine building block is allowed to react with a 2-halomethylpyri-
dine (method B). In the third route, condensation of a trans-2,5-
disubstituted pyrrolidine and a pyridinyl carboxylic acid or alde-
hyde followed by reduction forms the ligand (method C). All three
methods rely on the asymmetric synthesis of chiral 1,4-diols since
they are the precursors for the dimesylates of method A as well as
the pyrrolidines of methods B and C.
2: R = 4-t-BuPh
3: R = 2-naphthyl
4: R = 4-OMePh
5: R = mesityl
Scheme 2. Preparation of chiral 1,4-diaryl-1,4-diols.
demethylation of the two methoxy groups to form the diphenol.42
The phenol groups might be oxidized during the workup proce-
dure. Diketone 5 was found to be resistant to common reducing re-
agents, including LiAlH4, NaBH4, and LiBH4. Only starting material
was recovered. MM2* calculations (Chem3D 2000) of the most sta-
ble conformation of 5 revealed that the phenyl groups were per-
pendicular to the plane of the adjacent carbonyl groups. This
causes both faces of the carbonyl groups to be shielded by the
2,6-dimethyl groups on the phenyl rings, preventing the reducing
agent from approaching.
OMs
R'
R
+
NH2
n
R
Method A
Method B
N
OMs
R
R
2.3. Synthesis of pyrrolidines
R'
R'
+
+
N
n
X
n
HN
N
N
N
Once the chiral diols were prepared, they were transformed into
the corresponding dimesylates or pyrrolidines to use in method A
or method B, respectively. Pyrrolidines 1132 and 1233 were pre-
pared according to literature procedures. Chiral diols 6 and 7 were
first transformed into their corresponding dimesylates. We found
that the dimesylates of diols 6 and 7 are thermally labile and are
decomposed at ambient temperature even in solution. Even at
ꢀ20 °C, decomposition of the dimesylate of 6 in CH2Cl2 solution
could be observed. Therefore, it was impossible to isolate these
two dimesylates and characterize them spectroscopically. Instead,
a procedure that maintains both dimesylates at low temperature
and avoids total removal of solvent was used for the synthesis of
ligands using these two intermediates. The addition of allylamine
to the crude reaction mixture of the corresponding dimesylate
afforded the cyclized products 13 and 14 (Scheme 3). Similar to
the result reported by Chong,33 a small amount of the meso diaste-
reomer of the products was always observed, indicating a small de-
gree of degradation of stereochemistry. However, the rac and meso
diastereomers were easily separated by flash chromatography. The
N-allyl-protected trans-2,5-diarylpyrrolidine 14 was transformed
into the trans-2,5-diarylpyrrolidine 15 in 86% yield in the presence
of Wilkinson’s catalyst (Scheme 3). Due to the poor yield in the
preparation of 13 (7%), a ligand containing trans-2,5-ditolylpyrroli-
dine was prepared using the direct cyclization reaction (Method A)
R
R
R
n = 1, 2
R
Method C
R'
X
HN
O
(X = Cl, H)
Scheme 1. Retrosynthetic analysis of ligand synthesis.
2.2. Synthesis of chiral 1,4-diols
methodology developed by Masamune32 and
Adopting
a
Chong,33 the synthesis of trans-2,5-disubstituted pyrrolidine-con-
taining ligands started with the preparation of a series of chiral
1,4-diols. These were generally prepared from the corresponding
1,4-diketones by asymmetric reductions. Commercially available
2,5-hexanedione was reduced to (+)-(2S,5S)-2,5-hexanediol by Ba-
ker’s yeast using a slightly modified procedure reported by Lieser34
and Haufe.35 However, all other 1,4-diketones were 1,4-diaryl-1,4-
butanediones that were prepared by reduction of the correspond-
ing 1,4-diaryl-2-buten-1,4-diones, in turn were readily prepared
in good yields using the Friedel–Crafts reaction.36–40 Asymmetric
reduction of the 1,4-diaryl-1,4-butanediones to the corresponding
diols was achieved using (ꢀ)-DIP-Cl as the reducing agent,33 orig-
inally developed by Brown et al. (Scheme 2).41 Thus, diketones 1
and 2 were reduced selectively with (ꢀ)-DIP-Cl without difficulty
to provide chiral diols 6 and 7 in high diastereo- and enantioselec-
tivities. Diols 6 and 7 are colorless crystalline compounds that can
be easily purified by recrystallization from CH2Cl2–hexanes. Dike-
tone 3 was extremely difficult to reduce, with 3 equiv of (ꢀ)-DIP-
Cl per carbonyl group needed to drive the reaction to completion.
Although the selectivity was good (>99% ee, >98% de), the yield
was poor (19%, Scheme 2). The major product isolated was 2,5-
NH2
OH
MsCl, Et3N
OMs
R
R
R
-40 oC to RT
R
CH2Cl2
-40 oC
OH
OMs
6: R = p-Tol
7: R = 4-t-BuPh
R
Wilkinson's
catalyst
H
dinaphthyltetrahydrofuran, possibly formed via
intermediate.
a carbocation
N
R
R
N
R
84:16 MeCN/H2O
heat
Asymmetric reduction of diketones 4 and 5 proved not to be
feasible using (ꢀ)-DIP-Cl. The attempted reduction of 4 gave a pink
gel-like compound as the crude product after workup and the de-
sired product could not be isolated. The reaction was carried out
under strongly acidic conditions; a strong acid can promote the
15: R = 4-t-BuPh, 86%
13: R = p-Tol, 7% from 6
14: R = 4-t-BuPh, 35% from 7
Scheme 3. Synthesis of trans-2,5-diarylpyrrolidine.

1674
H. Chen et al. / Tetrahedron: Asymmetry 20 (2009) 1672–1682
of the corresponding dimesylate with a pyridinylmethyl amine
(vide infra).
NO2
OCH3
1. Ac2O, heat, 90%
2. NaOH, heat, 60%
OCH3
N
MeOH
H
N
H
N
3. PBr3, 83%
K2CO3
100%
R
N
O
R
N
O
Ph
Ph
Br
R
27: R = H
28: R = Me
11
12
20: R = H
21: R = Me
25: R = H
26: R = Me
Scheme 5. Synthesis of 2-bromomethyl-4-methoxypyridine and 2-bromomethyl-
4-methoxy-6-methylpyridine.
2.4. Synthesis of pyridines
To utilize method B, 2-bromomethylpyridine moieties were
needed to be prepared. 2-Picoline 16 and 2,6-dimethylpyridine
17 were oxidized (mCPBA) to afford the corresponding N-oxides
18 and 19, respectively.43 The N-oxides were then nitrated at the
4-position to give 2043 and 21.44 Subsequent transformation of
20 into an acetate by rearrangement in acetic anhydride at
100 °C45 yielded 22 in 90% yield (Scheme 4). After hydrolysis, the
hydroxymethyl pyridine 23 was further transformed to the desired
bromide 24 by treatment with PBr3. The nitro group at the 4-posi-
tion of 20 and 21 can be easily transformed into an electron-donat-
ing group, such as a methoxy group, to afford 2545 and 2646 in
quantitative yields (Scheme 5). After rearrangement in acetic anhy-
dride, hydrolysis, and bromination with PBr3, bromides 27 and 28
were obtained. Quinoline-based bromide 3147 was prepared from
2-quinolinecarboxylic acid 29 through esterification, reduction,
and bromination (Scheme 6).
O
OH
Br
OH
1. SOCl2
2. MeOH
48% HBr
80%
N
N
N
3. NaBH4
90%
30
31
29
Scheme 6. Synthesis of 2-bromomethylquinoline.
34–40 all contain a trans-2,5-diphenylpyrrolidine moiety, but are
differentiated by substituents on the pyridine ring or by the
lengths of the linkage between the pyridine and pyrrolidine moie-
ties. Ligands 41 and 42 have remote modification on the para-posi-
tion of the phenyl ring of the pyrrolidine moiety.
HNO3, H2SO4
heat, 90%
mCPBA
80%
CH3
Y
R
N
O
R
R
N
R''
N
N
n
N
16: R = H
17: Me
18: R = H
19: R = Me
CH3
R'
N
32: n = 1
33: n = 2
R
R
R' R"
Y
NO2
NO2
Ph
H
H
NO2
36:
37:
38:
39:
40:
Ph
Ac2O, heat
90%
NaOH
Ph
Ph
Ph
Ph
H
CH3
CH3
H
H
H
OCH3
H
OCH3
H
heat, 75%
N
n
R
N
O
OAc
N
N
Ph
benzo
22
20: R = H
21: R = Me
41: p-Tol
42: p-tBuPh
H
H
H
H
34: n = 1
35: n = 2
H
H
NO2
NO2
Both methods A and B worked satisfactorily to prepare the li-
gands. However, we encountered a difficulty in using method C.
In the case of using trans-2,5-diphenylpyrrolidine to construct
new ligands using method C with an acyl chloride, the first step
proceeded well to form the amide in high yield. However, reduc-
tion of the amide using LiAlH4 or NaBH4 did not occur. For this rea-
son, only the first two strategies were applied throughout the
remainder of this study.
PBr3, CH2Cl2
OH
Br
80%
N
N
23
24
Scheme 4. Synthesis of 2-bromomethyl-4-nitropyridine.
Ligands 32–35 were prepared according to method A (Scheme
7).16 Ligand 41, which is remotely modified on the two phenyl
rings on the pyrrolidine ring, was also prepared by method A. Li-
gand 38 was prepared either by direct cyclization (method A), or
by the coupling of trans-2,5-diphenylpyrrolidine with 6-methylpy-
ridinylmethyl bromide (method B, vide infra). All the other ligands
are prepared by method B.
The syntheses of 32–34 proceeded without significant epimer-
ization.16 There was also no detected erosion of stereochemistry
for the synthesis of ligand 35. However, degradation of stereo-
chemistry was observed in the preparation of ligands 38 and 41
by direct cyclization (method A). Following the same procedure
for the preparation of 34,16 synthesis of 38 gave a product, after
column chromatography, with a de of 76% (1H NMR), while synthe-
sis of 41 afforded only a mixture with a de of about 50% (deter-
mined by 1H NMR). Fortunately, only after one recrystallization
At ambient temperature, bromides 27 and 28, which bear an
electron-donating methoxy group at the 4-position of the pyridine
ring, rapidly undergo self-condensation to form yellow solids that
are insoluble in common organic solvents but are soluble in water.
Fortunately, decomposition of the bromides can be avoided when
the compounds are kept in solution. Thus, a procedure was devel-
oped to avoid the isolation of the bromides. For this reason, these
intermediate bromides were used without purification.
2.5. Synthesis of ligands
Eleven chiral pyridine–pyrrolidine ligands have been prepared
(Chart 3). Ligands 32 and 33 are constructed with trans-2,5-
dimethylpyrrolidine 11, and they differ in the length of the linkage
between the pyridine and the pyrrolidine moieties. All other li-
gands are constructed with trans-2,5-diarylpyrrolidines. Ligands

H. Chen et al. / Tetrahedron: Asymmetry 20 (2009) 1672–1682
1675
R
H
N
MsO
R
neat
R
Y
N
R
R
Y
+
R
N
n
R''
R'
NH2
n
oC
to RT
N
R''
R'
N
0
OMs
R
N
MeCN, K2CO3
40 o
Br
43: n = 1
44: n = 2
32: 54%
33: 68%
34: 75%
35: 53%
41: 50%
N
C
R
45: R = Me (ent)
46: R = Ph
47: R = p-Tol
36: 82%
37: 60%
38: 80%
39: 81%
40: 85%
42: 55%
46
Ph
neat
Scheme 8. Synthesis of chiral pyridine–pyrrolidine ligands by method B.
NH2
N
0
oC
to RT
80%
N
N
Ph
48
38
toluene, the reaction required 70 h to achieve total conversion with
an isolated yield of only 30%. The enantioselectivity was also poor
in non-polar solvents. Acetonitrile provided somewhat better
enantioselectivity for the reaction, but the yield was below 50%.
Dichloromethane appeared to be the solvent of choice for this reac-
tion. When the reaction was carried out in CH2Cl2, it proceeded in
69% yield with comparable enantioselectivity (Table 1, entries 1–
3). Therefore, most reactions carried out thereafter were in CH2Cl2.
Scheme 7. Synthesis of chiral pyridine–pyrrolidine ligands by method A.
from hexanes, the de and ee of 38 increased to >99% (by chiral
HPLC). Similarly, only one recrystallization of 41 from hexanes
brought the de to >98% and ee to >99%. The constitution and con-
figuration of 41 was confirmed by X-ray structure analysis (Fig. 1),
and therefore also confirmed the absolute configuration of the cor-
responding chiral diol 4. It is unlikely that epimerization occurs
during the mesylation step in the conversion of chiral alcohol
1,4-diphenyl-1,4-butanediol to the corresponding dimesylate 46,
since previously isolated samples of this intermediate showed
little to no degradation in diastereomeric purity. Furthermore,
cyclization of 46 with 2-(aminomethyl)pyridine, leading to ligand
34, proceeded with little epimerization.
Table 1
Palladium-catalyzed allylic alkylation of (rac)-(E)-1,3-diphenylprop-2-enyl acetate
[Pd(η3-C3H5)Cl]2
OAc
CH(COOMe)2
Ligand
CH2(COOMe)2
BSA, KOAc
RT
No.
Liganda
Solvent
Time (h)
Yield (%)
%eeb
Config.c
1
2
3
4
5
6
32
32
32
33
33
34
34
34
35
36
37
38
38
39
40
41
41
42
CH2Cl2
MeCN
Tol.
CH2Cl2
Toluene
CH2Cl2
THF
28
23
70
71
238
9
185
185
168
240
11
21
100
20
69
46
30
43
11
100
85
30
0
100
100
83
85
100
60
100
100
100
21
27
20
17
12
62
53
46
—
63
64
35
33
40
20
70
70
68
(S)
(S)
(S)
(R)
(R)
(R)
(R)
(R)
—
(R)
(R)
(S)
(S)
(S)
(S)
(R)
(R)
(R)
7d
8d,e
9
THF
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
10
11
12
13e
14
15
16
17e
18
115
10
26
Figure 1. ORTEP diagram of ligand 41 showing atom labeling scheme. Thermal
ellipsoids at 30% probability.
10
a
All reactions were run using 10 mol % of ligand and 2.5 mol % of Pd source
Method B was used to prepare ligands 36–40 and 42 (Scheme
8). A bromomethylpyridine and a pyrrolidine were allowed to react
in acetonitrile or acetone at 40 °C in the presence of K2CO3 to form
the corresponding pyridine–pyrrolidine ligand in good yield. Since
this method does not involve any reaction on the stereogenic cen-
ters, we did not observe any degradation of the pre-established
stereochemistry. In this respect, method B has the advantage over
method A for the synthesis of these ligands.
except as indicated.
b
The percentage ee was determined by 1H NMR (CDCl3) using the shift reagents
Pr(tfc)3 or Eu(hfc)3.
c
Determined by comparison of the specific rotation with the literature values.
NaH was used as the base instead of BSA + KOAc.
These reactions were run using 5 mol % of the ligands and 1.2 mol % of Pd
d
e
source.
Several trends have been identified from the results shown in
Table 1. First, the trans-2,5-diarylpyrrolidine moiety provides bet-
ter stereoselectivity in this catalytic system than the dimethyl ana-
logues. When 10 mol % of ligand and 2.5 mol % of palladium source
were used in combination with BSA as the base, trans-2,5-diphe-
nylpyrrolidine-containing ligand 34 catalyzed this reaction to com-
pletion in less than 10 h, affording 100% isolated yield with an ee of
62% (Table 1, entry 6), while the structurally similar trans-2,5-
dimethylpyrrolidine analogue only catalyzed this reaction to afford
up to 69% yield (Table 1, entry 1) and up to 27% ee (Table 1, entry
2). When the base was changed to NaH, and the solvent to THF, the
2.6. Catalytic results for palladium-catalyzed allylic alkylation
Upon preparation of these pyridine–pyrrolidine ligands, we
sought to test them in transition metal-catalyzed asymmetric reac-
tions. We applied these ligands in the palladium-catalyzed allylic
alkylation reaction of 1,3-diphenylprop-2-enyl acetate with di-
methyl malonate as a nucleophile (Table 1).
The effect of solvent on the reaction was first tested using li-
gand 32 in the presence of BSA and a catalytic amount of potas-
sium acetate. Chlorinated solvents resulted in higher yields. In

1676
H. Chen et al. / Tetrahedron: Asymmetry 20 (2009) 1672–1682
time of the reaction catalyzed by the palladium complex of 34 was
greatly increased while the stereoselectivity was decreased, result-
ing in a yield of only 85% after 185 h and an ee of only 53% (Table 1,
entry 7). When the loading of the catalyst was decreased, both the
yield and ee for the reaction involving 34 as catalyst precursor
were decreased, affording a yield of only 30% in 185 h and an ee
of 46% (Table 1, entry 8). The other trans-2,5-diarylpyrrolidine-con-
taining ligands also provided better yields and selectivities for this
reaction than the dimethyl analogues.
enantiomer as the major product in 70% ee and quantitative yield
within 10 h (Table 1, entry 16). Lowering the catalyst loading to
half only resulted in a longer reaction time, but the enantioselec-
tivity remained the same (Table 1, entry 17). Using the sterically
bulkier trans-2,5-bis(4-tert-butylphenyl)pyrrolidine moiety in li-
gand 42 did not provide better selectivity, but the reactivity re-
mained unchanged (Table 1, entry 18).
2.7. Structural studies of the
complexes
g3ꢀallyl palladium ligand
Changing the length of the linkage between the pyridine moiety
and the pyrrolidine moiety resulted in the stereoselectivity of the
reaction being reversed and the activity of the catalyst being de-
creased. For example, a palladium complex of ligand 32 catalyzed
the allylic alkylation reaction in CH2Cl2 to afford the product in
21% ee with the major product with an (S)-configuration (Table
1, entry 1). Ligand 33, which differs from 32 by an ethylene bridge
rather than a methylene bridge between the pyridine and the pyr-
rolidine moieties, afforded the (R)-isomer as the major product
with a lower ee and a lower yield (Table 1, entry 4). This reaction
also required a much longer time for completion (Table 1, entry
4 vs entry 1, entry 5 vs entry 3). When the ligands contain a
trans-2,5-diarylpyrrolidine moiety, increasing the length of the
bridge greatly decreased the reactivity of the ligand, resulting in
no detectable product after running the reaction in order for one
week (Table 1, entry 9 vs entry 6). These results suggest that for
better selectivity and reactivity in this catalytic system, the li-
gand–metal complex must have a five-membered rather than a
six-membered chelate ring. A similar inversion of enantioselectiv-
ity with a longer bridge was reported for pyridine–aziridine
ligands.29
It was found that modification on the 6-position of the pyridine
ring caused reversal of the stereoselectivity for the allylic alkyl-
ation, similar to an increase in the chelate size. Ligand 34 catalyzed
the reaction to afford the major product as the (R)-enantiomer.
However, ligand 38, which only differs from 34 in that it has a
methyl group on the 6-position of the pyridine ring, catalyzed
the reaction with the opposite stereoselectivity, giving the (S)-
enantiomer as the major product (Table 1, entries 12 and 13).
The same phenomenon was also observed when ligands 39 and
40 were used in the catalytic reaction (Table 1, entries 14 and
15). A similar inversion of enantioselectivity with a C6-pyridine
substituent was reported for pyridine–aziridine ligands.29
Changing the electronic properties of the substituents on the 4-
position of the pyridine ring has no significant effect on the stere-
oselectivity of the reaction. Instead, the electronic properties of the
substituents on this position only affected the reactivity of these li-
gands. We found that electronically deficient ligands had a much
lower reactivity than those of the electronically rich ones. Ligands
34, 36, and 37 afforded essentially the same selectivity for the cat-
alytic reaction. However, the reaction times were quite different.
The reaction catalyzed by 34 was completed in about 9 h with an
ee of 62% (Table 1, entry 6). With an electron-donating methoxy
group at the 4-position of the pyridine ring, ligand 37 also cata-
lyzed the allylic alkylation to completion in approximately the
same time with similar selectivity (Table 1, entry 11). However, li-
gand 36, which has a strong electron-withdrawing group on the 4-
position of the pyridine ring, catalyzed the reaction to completion
in 240 h (Table 1, entry 10). Ligands 38 and 39, which only differ
from each other by an additional methoxy group at the 4-position
of the pyridine ring, gave comparable selectivities and reactivities
(Table 1, entries 12 and 14).
In an attempt to better understand the role of these ligands in
the palladium-catalyzed allylic alkylation reaction and to help
optimize the system, consideration of the structures of the
palladium ligand complexes was appropriate. For a C2-symmetric
ligand and a mirror symmetric allyl group, the
3-allyl metal li-
g
3-allyl
g
gand complex has only one single configuration. Therefore the ste-
reoselectivity of the reaction depends solely on the regioselectivity
of the nucleophilic attack on the
the ligand lacks C2-symmetry, the configuration of the
group (endo or exo) makes a difference. Complexes A and B differ
in the configuration of the
3-allyl moiety (Scheme 9). In this sys-
g
3-allyl moiety.48 However, when
g
3-allyl
g
tem, nucleophilic attack can occur at any of the 4 allylic termini.
For the reaction to exhibit stereoselectivity, there should not only
be discrimination between the two configurations of the complex
at the transition state, but also a bias between the reactivity of
C1 and C3. The allyl carbon trans to pyrrolidine will be commonly
referred to as C1, and that trans to pyridine as C3. Hence, an ideal
ligand directs nucleophilic attack to only one of the four positions.
H
Ph
Ph
1
1
Nu
2
3
N
N
1
3
Pd
Nu
Nu
Pd
Nu
N
N
4
3
H
Ph
Ph
A: endo
B: exo
Scheme 9. Possible nucleophilic attacks on the
plexes of non-C2-symmetric ligand 34.
g
3-allyl palladium ligand com-
A previously obtained crystal structure of the allyl palladium
complex of 32 exhibited a disordered allyl group.16 This strongly
suggests that trans-2,5-dimethylpyrrolidine is not effective at
restricting the conformation of the allyl group. We have obtained
an X-ray structure of the allyl palladium complex of ligand 34
(49, Fig. 2). Ligand 34 was allowed to react with the allyl palladium
chloride dimer and NaClO4 to form the complex. The crystals of the
complex were grown from CH2Cl2–ether solution. The diastereo-
mer with the allyl group at the endo position was found in the crys-
tal. One of the phenyl groups of the pyrrolidine moiety masks the
allyl C3 position from the top face. As expected, the structure
shows a bias for the two Pd–C bonds. Pd–C3 is longer than Pd–
C1 by about 0.03 Å. This suggests that the atom trans to C3 is a bet-
ter p
-acceptor (pyridine sp2-nitrogen). The pyridine sp2-nitrogen is
softer than the sp3-nitrogen, and therefore provides a stronger
‘trans-influence’.49 That the Pd–C3 bond is longer than the Pd–C1
bond also suggests that the Pd–C3 bond is weaker and should be
broken more easily than Pd–C1 bond when subjected to nucleo-
philic attack.48 Based on the discussion above, we believed that
the nucleophilic addition on the allyl group is more likely to hap-
pen at the C3 position (trans to pyridine) in this system.
Remote modification of the pyrrolidine ring provided enhanced
selectivity and reactivity. Ligands 41 and 42 exhibited a moderate
improvement over ligand 34 and are the most promising ligands in
this series to date. When trans-2,5-ditolylpyrrolidine-containing li-
gand 41 was used as a catalyst precursor, the reaction gave the (R)-
2.8. Molecular mechanics (MM3*) study of the
palladium complexes
g
3-allyl
Understanding the origin of the stereoselectivity of the catalytic
processes can aid the optimization of the ligands for better

H. Chen et al. / Tetrahedron: Asymmetry 20 (2009) 1672–1682
1677
Ph
Ph
Ph
N
Ph
Pd
N
Ph
N
Pd
Ph
N
N
N
Ph
Ph
Ph
Pd
endo
exo
Ph
2.104 Å
2.131 Å
Global minimum (kJ/mol) Energy diff. (kJ/mol) Product
Config.
ligand
endo
279.93
309.00
exo
exo-endo
N
Ph
34
38
281.37
321.82
–1.44
(R)
(S)
Pd
N
–12.82
Ph
Ph
N
Ph
N
N
Pd
N
g
3-allyl)ꢂPdꢂ34]+ClO4 49 with thermal ellipsoids
ꢀ
Figure 2. (l.) ORTEP diagram of [(
Pd
Ph
2.192Å
Ph
at 30% probability. The perchlorate counterion and the hydrogen atoms, except for
those bonded to the chiral carbon atoms, are omitted for clarity. (r.) Two alternate
views of 49.
2.129Å
Ph
2.133Å
2.195Å
Ph
Ph
1
Ph
3
1
3
(η3-1,3-diphenylallyl)Pd-34
Ph
selectivity and activity. We used computational chemistry to ob-
tain energy-minimized structures of the
g
3-1,3-diphenylallyl Pd
Ph
complexes of ligands 34 and 38 using Norrby’s parameters for
the MM3* force field for palladium-catalyzed allylic alkylation
(MacroModel 6.0).50–52 To facilitate finding global minimum struc-
tures, the calculation was divided into two subsets, one with the
N
N
N
N
Pd
Pd
Ph
2.188Å
Ph
2.159Å
2.154Å
2.179Å
g
3-1,3-diphenylallyl moiety having an endo configuration and
Ph
Ph
Ph
1
one with the exo configuration. Monte Carlo conformational
searches were performed to find multiple conformations of the
two configurations of each complex and then the respective global
minima. The calculated energies for the global minimum of both
endo and exo configurations were compared, giving quantitative re-
sults of which diastereomer was predicted to be more abundant
(Fig. 3). Monte Carlo searches found multiple conformations for
Ph
3
1
3
3-1,3-diphenylallyl)Pd-38
(η
Figure 3. Calculated Pd–C bond lengths for the complexes of ligands 34 and 38.
ee, while 34 provided up to 65% ee, this is also not likely to be
the case. If an attack always occurs at the major ground state dia-
stereomer at the same position (i.e., trans to the pyridine or pyrrol-
idine nitrogen), both should form the same major product. This
prediction then conflicts with our experimental data [34 gave (R)
and 38 gave (S)].
We propose that nucleophilic attack occurs mainly at the less
stable exo diastereomer at the terminus trans to the pyridine rings
(C3, Scheme 10). The less stable diastereomers (exo) are more reac-
tive toward nucleophilic attack since their energies are higher than
the more stable diastereomers (endo). In the case of ligand 34, the
difference between the two Pd–C bonds is large (ꢁ0.06 Å) hence
we can consider the carbon trans to the pyridine nitrogen (sp2)
both configurations of each
g
3-1,3-diphenylallyl Pd complex, indi-
cating the flexibility of the pyrrolidine ring. Since ligand flexibility
can decrease the stereoselectivity of the catalytic reactions by
increasing the number of accessible transition states, the flexibility
of the pyrrolidine ring may be responsible for the moderate stere-
oselectivity exhibited by these pyridine–pyrrolidine ligands.26
The calculated distance between the palladium and the allyl
terminus trans to the pyridine ring was consistently longer than
that between the palladium and the allyl terminus trans to the pyr-
rolidine ring (Fig. 3). This supports our assumption based on the
trans influence. Since the pyridine nitrogen is also a p-acceptor, it
weakens the Pd–C bond trans to it. This allows the Pd–C bond to
be broken more easily. Koga48 and Pfaltz53 in separate studies ob-
tained evidence that the allyl terminus with a longer Pd–C bond
possessed more positive charge and therefore was more readily at-
tacked by nucleophiles.
1
Ph
3
N
Ph
N
Ph
Nu
Ph
Pd
S
1
Pd
Ph
N
N
Nu
3
R
Ph
slightly more stable
(η3-1,3-diphenylallyl)Pd-34
Ph
Ph
2.9. Mechanistic consideration for the palladium-catalyzed
allylic alkylation reaction
Since the major product from both catalysts [(R) for 34 and (S)
for 38] is known, a brief discussion regarding the origin of the ste-
reoselectivity is possible. If the nucleophilic attack always occurs
trans to the pyrrolidine nitrogen, then in the case of 34, attack oc-
curs at the more stable endo diastereomer, while in the case of 38,
the attack occurs at the less stable exo diastereomer to provide the
observed stereoselectivity. This is unlikely to be true. If an attack
always occurs trans to the pyridine nitrogen on the more stable
diastereomer, then the selectivity for ligand 38 should be better
than that for 34. Taking into account that 38 only provided 33%
1
Ph
3
3
N
Ph
N
Ph
Ph
Nu
Pd
1
Pd
Ph
S
N
N
Nu
Ph
R
Ph
Ph
more stable
(η3-1,3-diphenylallyl)Pd-38
Scheme 10. Proposed mechanism for the palladium-catalyzed allylic alkylation.

1678
H. Chen et al. / Tetrahedron: Asymmetry 20 (2009) 1672–1682
bearing more positive charge than the carbon trans to the pyrroli-
dine nitrogen (sp3), and therefore is more subject to nucleophilic
attack.48,53 Because in the ground state the complex of ligand 34
exists in about a 1:1 ratio for its two configurations (based on 1H
NMR studies not shown), the selectivity results mainly from the
electronic difference between the two allylic termini.
(ꢀ)-DIP-Cl (4.18 g, 13.0 mmol) at ꢀ78 °C under N2. The mixture
was stirred at ꢀ78 °C for 2 h and then allowed to warm to rt over
a 24-h period. The solvent was removed in vacuo, and the residue
was stirred at 40 °C in vacuo (0.3 torr) for 24 h. The residue was
dissolved in dry ether (50 mL) and cooled to 0 °C, and diethanola-
mine (2.30 g, 21.9 mmol) was added. The mixture was stirred vig-
orously at 0 °C for 30 min and at rt for 24 h. The resulting white
precipitate was removed by filtration through a pad of Celite. Sol-
vent was removed in vacuo and the residue was dissolved in hex-
ane (30 mL) and stored in a freezer overnight. The colorless
precipitate was collected by filtration and recrystallized from hex-
anes–CH2Cl2 to afford 6 as a colorless solid (1.01 g in two crops,
83%). This product can be used directly in the next step. For analyt-
ical purpose, a small portion of this product was further purified by
recrystallization in hexane–ethanol to afford the product as color-
less needles: 1H NMR (400 MHz, CDCl3) d 7.20 (d, J = 8.1 Hz, 4H),
7.12 (d, J = 8.1 Hz, 4H), 4.66 (t, J = 5.6 Hz, 2H), 2.41 (br s, 2H),
2.32 (s, 6H), 2.19–1.87 (m, 4H), 1.80–1.73 (m 4H); 13C NMR
(100 MHz, CDCl3) d 142.6, 138.0, 130.0, 126.7, 75.3, 36.8, 22.0;
In the case of ligand 38, the distance between the two Pd–C
bonds is much smaller (0.02–0.03 Å). The difference of the elec-
tronic properties of both allylic termini still favors nucleophilic at-
tack on the carbon trans to the pyridine nitrogen. However, since
the major diastereomer (endo) is much more stable than the minor
diastereomer (exo, favored by 12.82 kJ/mol), the endo diastereomer
is more abundant than the exo diastereomer in solution (1H NMR
studies). Nucleophilic attack on the more abundant endo diastereo-
mer at the allylic terminus trans to the pyridine nitrogen contrib-
utes to the formation of the (S)-enantiomer of the product. The
less effective results (35% ee) and the opposite selectivity (S vs R)
obtained from the reaction involving ligand 38 result from the
competition of the aforementioned two reaction routes. However,
it is difficult to rule out other possibilities for the origin of stereose-
lectivity, including the involvement of syn–anti allyl isomers, as
well as the unfavorable steric interaction between developing
product and the pyridine methyl group 38 in a late transition
state,54 at this time.
½
a 2D5
ꢃ
¼ ꢀ47:0 (c 0.65, CH2Cl2); mp = 118–119 °C; MS (FAB+) m/z
(rel intensity) 253 ([MHꢀH2O]+, 100), 235 (52), 185 (92), 137
(38), 93 (97); exact mass (FABꢀ) m/z calcd for C18H21O2 ([MꢀH]ꢀ)
269.1542, found 269.1548; ee = 99.5%, de = 98.0% [Chiralcel OD,
99.5:0.5 hexanes/ethanol, 0.5 mL/min, 65.4 min (S,S), 67.6 min
(meso), 70.3 min (R,R)]. Anal. Calcd for C18H22O2 (270.2): C, 79.96,
H, 8.20. Found: C, 79.74, H, 8.17.
3. Conclusions
4.2.2. (ꢀ)-(1S,4S)-1,4-Di(4-tert-butylphenyl)-1,4-butandiol 7
Freshly distilled THF (160 mL) was added to a mixture of 1,4-
bis(4-tert-butylphenyl)-1,4-butandione (2, 8.00 g, 23 mmol) and
(ꢀ)-DIP-Cl (16.00 g, 50.0 mmol) at ꢀ78 °C under N2. The mixture
was stirred at ꢀ78 °C for 2 h and then allowed to warm to rt over
a 24-h period. The solvent was removed in vacuo, and the residue
was stirred at 40 °C in vacuo (0.3 Torr) for 24 h. The residue was
dissolved in dry ether (350 mL) and cooled to 0 °C, and diethanol-
amine (8.00 g, 76 mmol) was added. The mixture was stirred vigor-
ously at 0 °C for 30 min and at rt for 24 h. The resulting white
precipitate was removed by filtration through a pad of Celite. The
solvent was removed in vacuo and the residue was purified by
flash chromatography (SiO2, hexanes/ether = 1:1) to afford 7 as a
colorless solid (5.02 g, 63%). 1H NMR (250 MHz, CDCl3) d 7.35 (d,
J = 6.5 Hz, 4H), 7.25 (d, J = 6.5 Hz, 4H), 4.66 (m, 2H), 2.77 (broad,
2H), 1.88 (m, 4H), 1.30 (s, 18H); C NMR (62.896 MHz, CDCl3) d
In conclusion, we have successfully prepared a novel type of
chiral bidentate ligands based on trans-2,5-disubstituted pyrroli-
dine building blocks. These ligands are effective catalyst precursors
for the palladium-catalyzed allylic alkylation and provide moder-
ate to good stereoselectivity. The stereoselectivity of this reaction
catalyzed by the palladium complexes of the pyridine–pyrrolidine
ligands could be adjusted by changing the substituent at the 6-po-
sition of the pyridine ring, or the substituents at the pyrrolidine
ring as well as the chelate size of the complexes. MM3* calcula-
tions, in concert with the catalytic results, suggest that nucleo-
philic attack on the
g
3-1,3-diphenylallyl moiety predominantly
occurred trans to the pyridine ring at the less stable configuration
in the transition state, although other possibilities have not been
ruled out.55
150.4, 141.6, 125.6, 125.3, 74.4, 35.8, 34.5, 31.3; ½a D25
¼ ꢀ31:3 (c
ꢃ
4. Experimental
1.14, CH2Cl2); mp = 138–139 °C; MS (FAB+) m/z (rel intensity)
709.6 ([2M+H]+, 15), 335.3 (60), 263.2 (43), 207.2 (32), 185.1
(100), 93.2 (82), 57.4 (51); exact mass (FABꢀ) m/z calcd for
C24H33O2 ([MꢀH]ꢀ) 353.2481, found 353.2477; ee >95%, de >95%
(1H NMR, Mosher’s acid as chiral-shifting agent, no signal for the
stereoisomers could be found). Anal. Calcd for C24H34O2 (354.3):
C, 81.31, H, 9.67. Found: C, 81.02, H, 9.60
4.1. General methods
All reactions involving air- or moisture-sensitive reagents were
carried out under a nitrogen atmosphere using oven-dried glass-
ware. THF was either distilled from Na or dried over freshly acti-
vated 4 Å molecular sieves. CH2Cl2 was either distilled from CaH2
or dried over freshly activated 4 Å molecular sieves. Toluene was
distilled from Na. Pyridine and methanol were dried over freshly
activated 4 Å molecular sieves before use. Compounds 11,32 12,33
18–20,43 21,44 25,45 26,46 and 32–3416 were prepared according
to literature procedures. All other solvents and reagents were pur-
chased from commercial sources and used as received. NMR spec-
tra were recorded in CDCl3 solution with the solvent peaks (1H
7.24 ppm or 13C 77.0 ppm) as internal standards. Melting points
were not corrected.
4.2.3. (ꢀ)-(1S,4S)-1,4-Di(naphth-1-yl)-1,4-butandiol 8
Freshly distilled THF (20 mL) was slowly added to a mixture of
1,4-bis(1-naphthyl)-1,4-butandione (3, 0.2791 g, 0.80 mmol) and
(ꢀ)-DIP-Cl (2.5828 g, 8.1 mmol) at ꢀ78 °C. The solution was stirred
for 2 h at ꢀ78 °C and then for 12 h at rt. The solvent was removed
in vacuo and the residue was stirred in vacuo (0.1 Torr) at 40 °C for
24 h. Dry THF (20 mL) and diethanolamine (0.97 g, 16 mmol) were
then added at 0 °C and the mixture was stirred vigorously for
30 min at 0 °C and for 12 h at rt. After workup, a viscous oil was ob-
tained. Purification by flash chromatography (SiO2, hexane/
ether = 1:1) afforded 8 as colorless crystals. (50.5 mg, 19%): 1H
4.2. Synthesis of butandiols
4.2.1. (ꢀ)-(1S,4S)-1,4-Di(4-methylphenyl)-1,4-butandiol 6
Freshly distilled THF (35 mL) was added to a mixture of 1,4-
bis(4-methylphenyl)-1,4-butandione 1 (1.20 g, 4.50 mmol) and
NMR (400 MHz, CDCl3)
J = 3.5 Hz, 2H), 7.75 (d, J = 8.2 Hz, 2H), 7.67 (d, J = 7.1 Hz, 2H),
7.49–7.43 (m, 6H), 5.56 (s, 2H), 2.90 (d, J = 18.8 Hz, 2H), 2.23–
d 8.05 (d, J = 7.6 Hz, 2H), 7.82 (t,

H. Chen et al. / Tetrahedron: Asymmetry 20 (2009) 1672–1682
1679
2.21 (m, 2H), 2.16–2.11 (m, 2H); 13C NMR (100 MHz, CDCl3) d
to rt, it was poured into crushed ice (200 g) and the resulting mix-
ture was neutralized with 4 M NaOH and extracted with chloro-
form (3 ꢄ 50 mL). The combined chloroform layer was washed
with brine (2 ꢄ 50 mL), dried (MgSO4), and concentrated in vacuo
to afford a light brown oil. Flash chromatography (SiO2, hexane/
ether = 1:1) of the oil afforded 22 as an orange oil (10.2 g, 90%):
1H NMR (500 MHz, CDCl3) d 8.87 (d, J = 4.1 Hz, 1H), 8.05 (dd,
J = 1.9, 7.6 Hz, 1H), 7.94 (ddd, J = 7.6, 4.1, 1.9 Hz, 1H), 5.33 (s, 2
H), 2.20 (s, 3H).
141.1, 134.7, 131.2, 129.8, 128.8, 126.9, 126.4, 126.3, 123.9,
123.6, 72.2, 36.1; mp 145–147 °C; ½a D25
¼ ꢀ68:0 (c 1.1, CH2Cl2);
ꢃ
ee = 99.3%, de = 99.4% [Chiralcel OD, 90:10 hexanes/ethanol,
1 mL/min, 7.1 min (S,S), 13.3 min (meso), 21.9 min (R,R)]. Anal.
Calcd for C24H22O2 (270.2): C, 84.18, H, 6.48. Found: C, 84.14, H,
6.44.
4.3. Synthesis of pyrrolidines
4.3.1. (+)-(2R,5R)-1-Allyl-2,5-di-(4-tert-butylphenyl)pyrrolidine 14
To a cold solution (ꢀ78 °C) of methanesulfonyl chloride (2.0 mL,
27 mmol) in CH2Cl2 (100 mL) was added a solution of 5 (3.04 g,
8.6 mmol) and triethylamine (4.1 mL, 30 mmol) in CH2Cl2
(100 mL). The solution was stirred at ꢀ78 °C for 2 h and allylamine
(60 mL) was added. The mixture was allowed to gradually warm to
rt and stirred for 2 d. The mixture was concentrated in vacuo, and
the residue was diluted with ether (300 mL), washed with satd
NaHCO3 (2 ꢄ 30 mL), brine (2 ꢄ 30 mL), dried (MgSO4), and con-
centrated in vacuo to afford a yellow oil. Purification of this oil
by flash chromatography (SiO2, hexanes/ether = 30:1) afforded 14
as a colorless solid (1.50 g, 47%): 1H NMR (600 MHz, CDCl3) d
7.33 (d, J = 6.6 Hz, 4H), 7.25 (d, J = 6.6 Hz, 4H), 5.67 (m, 2H), 4.94
(m, 2H), 4.28 (t, J = 4.8 Hz, 2H), 2.96 (ddt, J = 1.8, 7.2, 18.0 Hz,
1H), 2.66 (dd, J = 7.2, 18.0 Hz, 1H), 2.48 (m, 2H), 1.90 (m, 2H),
1.32 (s, 18H); 13C NMR (125 MHz, CDCl3) d 149.5, 140.9, 137.1,
127.6, 124.9, 115.4, 65.0, 49.8, 34.4, 33.1, 31.4; mp 57–59 °C;
4.4.2. 2-Hydroxymethyl-4-nitropyridine 2343
2-Acetoxymethyl-4-nitropyridine 20 (1.39 g, 7.1 mmol) was
stirred with methanol (20 mL) and 1 M NaOH (9 mL) under N2
for 90 min. The mixture first became dark green and rapidly turned
to orange. Water (50 mL) was added and the resulting mixture was
extracted with ethyl acetate (3 ꢄ 50 mL). The combined organic
layer was dried (MgSO4), filtered, and concentrated in vacuo to af-
ford an orange oil which, after flash chromatography (SiO2, hexane/
ether = 4:1), afforded 23 as a pale yellow solid (0.86 g, 79%): 1H
NMR (250 MHz, CDCl3)
d 8.85 (d, J = 5.4 Hz, 1H), 8.05 (d,
J = 2.1 Hz, 1H), 7.92 (dd, J = 2.1, 5.4 Hz, 1H), 4.91 (d, J = 5.4 Hz,
2H), 3.19 (t, J = 5.4 Hz, 1H); 13C NMR (62.9 MHz, CDCl3) d 174.2,
163.3, 151.0, 114.9, 113.2, 64.4.
4.5. General procedure for the synthesis of ligands using
method A
½
a 2D5
ꢃ
¼ þ111:5 (c 1.1, CH2Cl2); MS (FAB+) m/z (rel intensity) 376.3
Dimesylate (1 equiv) and the corresponding amine (5 equiv)
were added together at 0 °C under N2 atmosphere. The mixture
was then allowed to warm slowly to rt and react for 96 h, after
which time the mixture was transferred to ice-cold 2 M NaOH
and the resulting solution was extracted with hexane (four times).
The combined hexane layers were dried (KOH pellets or MgSO4).
After filtration, the solvent was removed in vacuo and the resulting
oil was purified by flash chromatography (SiO2, hexanes/ether) to
afford the desired compound.
(MH+, 63), 246.1 (11). 185.1 (100), 93.2 (83); exact mass (FAB+)
m/z calcd for C27H38N (MH+) 376.3004, found 376.2997. Anal. Calcd
for C27H37N (375.3): C, 86.34, H, 9.93, N, 3.73. Found: C, 86.30, H,
9.89, N, 3.70.
4.3.2. (+)-(2R,5R)-2,5-Di-(4-tert-butylphenyl)pyrrolidine 15
(+)-(2R,5R)-1-Allyl-2,5-di-(4-tert-butylphenyl)pyrrolidine (14,
1.00 g, 2.7 mmol), (Ph3P)3RhCl (Wilkinson’s catalyst, 0.150 g,
0.174 mmol), and 100 mL of 84:16 acetonitrile/water mixture were
placed in a 200-mL round-bottomed flask. The flask was evacuated
and charged with argon three times, and then heated at reflux for
3 h before it was allowed to cool to rt. The mixture was diluted
with ether (200 mL), the organic layer was separated and washed
with brine (2 ꢄ 50 mL). The combined aqueous layers were back-
extracted with ether (50 mL). The combined organic layers were
dried (MgSO4) and concentrated in vacuo to afford a yellow solid.
Purification of the crude product by flash chromatography (SiO2,
hexanes/ether = 1:1) afforded 15 as an off-white solid (0.767 g,
86%): 1H NMR (500 MHz, CDCl3) d 7.36 (d, J = 8.5 Hz, 4H), 7.33 (d,
J = 8.5 Hz, 4H), 4.51 (t, J = 7.0 Hz, 2H), 2.37 (m, 4H), 1.96 (br s,
1H), 1.91 (m, 4H), 1.34 (s, 18H); 13C NMR (125 MHz, CDCl3) d
149.6, 142.9, 126.0, 125.3, 61.8, 35.4, 34.4, 31.4; mp 122–123 °C;
4.6. General procedure for the synthesis of ligands using
method B
A trans-2,5-disubstituted pyrrolidine (1 equiv), K2CO3 (2 equiv),
and a bromide (1 equiv) were allowed to react in MeCN under N2 at
40 °C for 24 h. The solid was removed by filtration and the filtrate
was concentrated in vacuo. The resulting oil was taken up in ether,
washed with satd aq NaHCO3 solution (twice) and brine (once),
and dried (MgSO4). After filtration, the solvent was removed in va-
cuo and the resulting oil was purified by flash chromatography
(SiO2, hexanes/ether) to afford the desired product.
4.7. Synthesis of pyrrolidine–pyridine ligands
½
a 2D5
ꢃ
¼ þ106:5 (c 1.13, CH2Cl2); MS (FAB+) m/z (rel intensity)
336.3 (MH+, 22), 246.1 (12). 185.1 (100), 93.2 (67); exact mass
(FAB+) m/z calcd for C24H34N (MH+) 336.2691, found 336.2691;
ee >99%, de >99% [Chiralcel OD, 99:1:0.1 hexanes/ethanol/triethyl-
amine, 1 mL/min, 15.6 min (R,R), 16.0 min (meso), 16.5 min (S,S)].
Anal. Calcd for C24H33N (375.3): C, 85.91, H, 9.91, N, 4.17. Found:
C, 85.80, H, 9.89, N, 4.09).
4.7.1. (+)-2-[(2R,5R)-2,5-Diphenylpyrrolidin-1-yl]methylpyridine
3416
(Method
A)
(ꢀ)-(1S,4S)-1,4-Bis(methanesulfonyloxy)-1,4-
diphenylbutane33 (1.51 g, 6.2 mmol) was prepared by the litera-
ture procedure. It was cooled to 0 °C and allowed to react with
2-aminomethylpyridine (3.50 g, 32.4 mmol) to afford ligand 34,
after flash chromatography (SiO2, hexane/ether = 1:1), as a pale
yellow solid (0.87 g, 44%): 1H NMR (400 MHz, CDCl3) d 8.45 (ddd,
J = 0.9, 1.8, 4.9 Hz, 1H), 7.59 (dt, J = 1.8, 7.7 Hz, 1H), 7.42 (d,
J = 7.6 Hz, 1H), 7.32–7.20 (m, 10H), 7.07 (m, 1H), 4.37 (br t,
J = 4.4 Hz, 2H), 3.66 (d, J = 15.7 Hz, 1H), 3.39 (d, J = 15.7 Hz, 1H),
2.60 (m, 2H), 2.04 (m, 2H); 13C NMR (100 MHz, CDCl3) d 160.3,
148.6, 143.3, 136.0, 128.2, 128.0, 127.0, 122.4, 121.2, 65.7, 53.3,
4.4. Synthesis of pyridines
4.4.1. 2-Acetoxymethyl-4-nitropyridine 2243
4-Nitro-2-picoline-N-oxide (20, 8.1 g, 58 mmol) was added to
hot (100 °C) acetic anhydride (30 mL) over a period of 3 min. The
mixture was maintained at 100 °C for 1 h until gas evolution
ceased. Ethanol (30 mL) was then cautiously added and the mix-
ture was allowed to reflux for 10 min. After the mixture was cooled
33.3; ½a 2D5
¼ þ131:6 (c 0.08, CHCl3); exact mass (EI) m/z calcd for
ꢃ
C23H25N2 (MH+) 315.1861, found 315.1869; de >98% (1H NMR),

1680
H. Chen et al. / Tetrahedron: Asymmetry 20 (2009) 1672–1682
ee >99% [Chiralcel OD, hexanes/isopropanol/triethylamine =
99:1:0.1, 0.5 mL/min, 19.2 min (R,R), 19.8 min (meso), 20.3 min
(S,S)].
7.30–6.95 (m, 10H), 6.94 (d, J = 2.5 Hz, 1H), 6.58 (dd, J = 3.0,
6.0 Hz, 1H), 4.37 (br dd, J = 4.0, 5.5 Hz, 2H), 3.82 (s, 3H), 3.59 (d,
J = 15.5 Hz, 1H), 3.36 (d, J = 15.5 Hz, 1H), 2.58 (m, 2H), 2.04 (m,
2H); 13C NMR (62.9 MHz, CDCl3) d 162.2, 150.0, 143.3, 128.2,
128.0, 127.0, 125.4, 108.0, 107.4, 65.7, 54.9, 53.2, 33.2;
4.7.2. (+)-2-[(2R,5R)-2,5-Diphenylpyrrolidin-1-yl]ethylpyridine 35
(Method A) 2-(2-Aminoethyl)pyridine (0.50 g, 4.1 mmol) was
allowed to react with (ꢀ)-(1S,4S)-1,4-bis(methanesulfonyloxy)-
1,4-diphenylbutane (0.50 g, 1.3 mmol) in the presence of triethyl-
amine (1.0 mL, 7.1 mmol). After aqueous workup as described be-
fore the crude product was purified by flash chromatography (SiO2,
hexane/ether = 2:1) to afford 35 as a colorless oil (0.22 g, 53%): 1H
NMR (500 MHz, CDCl3) d 8.40 (ddd, J = 0.5, 1.5, 5.0 Hz, 1H), 7.44 (dt,
J = 2.0, 7.5 Hz, 1H), 7.31–7.20 (m, 10H), 7.02 (ddd, J = 1.0, 4.5,
7.0 Hz, 1H), 6.82 (d, J = 8.0 Hz, 1H), 4.41 (m, 2H), 2.77–2.64 (m,
4H), 2.50 (m, 2H), 1.84 (m, 2H); 13C NMR (62.90 MHz, CDCl3) d
160.9, 148.9, 144.7, 135.7, 128.2, 127.7, 126.7, 123.0, 120.8, 66.2,
½
a 2D3
ꢃ
¼ þ81:7 (c 1.0, CH2Cl2); MS (FAB+) m/z (rel intensity) 345.3
(MH+, 55), 246.1 (9). 185.1 (100), 93.2 (80); exact mass (FAB+) m/
z calcd for C23H25N2O (MH+) 345.1967, found 345.197; de >98%
(1H NMR), ee >98% (1H NMR).
4.7.5. (+)-2-[(2R,5R)-2,5-Diphenylpyrrolidin-1-ylmethyl]-6-meth-
ylpyridine 38
(Method
A)
(1S,4S)-1,4-diphenyl-1,4-butandiol
(0.50 g,
2.1 mmol) was transformed to (ꢀ)-(1S,4S)-1,4-bis(methanesulfo-
nyloxy)-1,4-diphenylbutane33 by the literature procedure. It was
cooled to 0 °C and allowed to react with 2-aminomethyl-6-methyl-
pyridine (48, 1.81 g, 14.8 mmol) for 2 h and then at rt for 6 d, fol-
lowed by flash chromatography (SiO2, hexane/ether = 1:1) to give a
pale yellow oil, which solidified quickly in a freezer (0.49 g,
76.9%,%de = 75.5). Recrystallization of this product from hexane
afforded 38 as colorless crystals. (0.100 g, 15.8%). Compound 38
obtained in this way was identical with a sample obtained by
method B: 1H NMR (400 MHz, CDCl3) d 7.47 (dd, J = 7.4, 7.6 Hz,
1H), 7.29 (d, J = 7.6 Hz, 1H), 7.23 (m, 10H), 6.87 (d, J = 7.4 Hz, 1H),
4.35 (t, J = 4.2 Hz, 2H), 3.60 (d, J = 16.2 Hz, 1H), 3.38 (d,
J = 16.2 Hz, 1H), 2.55 (m, 2H), 2.37 (s, 3H), 1.99 (m, 2H); 13C NMR
(100 MHz, CDCl3) d 160.7, 157.9, 144.4, 137.1, 129.1, 128.8,
127.8, 121.5, 119.9, 66.6, 54.2, 34.2, 27.8; mp = 57–58 °C;
47.4, 37.5, 33.5; ½a D25
ꢃ
¼ þ116:2 (c 0.9, CH2Cl2); MS (FAB+) m/z (rel
intensity) 329.3 (MH+, 100), 236.2 (13), 185.2 (42), 93.2 (38); exact
mass (FAB+) m/z calcd for C23H25N2 (MH+) 329.2018, found
329.2022; de >98% (1H NMR).
4.7.3. (+)-2-[(2R,5R)-2,5-Diphenylpyrrolidin-1-yl]methyl-4-nitro-
pyridine 36
(Method B) 2-Hydroxymethyl-4-nitropyridine 23 (0.46 g,
3.0 mmol) was dissolved in dry CH2Cl2 (50 mL) and cooled to
ꢀ20 °C. Next, PBr3 (1.03 g, 3.77 mmol) was added and the mixture
was brought to reflux at 40 °C for 1 h. After the mixture was al-
lowed to cool to rt, it was neutralized with 30% aqueous ammonia,
and extracted with CH2Cl2 (2 ꢄ 15 mL), dried (MgSO4), and concen-
trated in vacuo to afford a yellow oil. Flash chromatography (SiO2,
hexane/ether = 1:1) of this oil afforded 24 as a pale yellow oil
(0.49 g, 76%): 1H NMR (250 MHz, CDCl3) d 8.85 (d, J = 5.5 Hz, 1H),
8.16 (d, J = 1.8 Hz, 1H), 7.92 (dd, J = 1.8, 5.5 Hz, 1H), 4.63 (s, 2H).
2-Bromomethyl-4-nitropyridine 24 thus obtained was allowed to
react with 12 (0.50 g, 2.2 mmol) and K2CO3 (0.50 g, 3.6 mmol) in
MeCN (20 mL) which, after workup and flash chromatography
(SiO2, CH2Cl2), afforded 36 as a thick yellow oil (0.66 g, 82%): 1H
½
a 2D3
ꢃ
¼ þ132 (c 1.15, CHCl3); MS (FAB+) m/z (rel intensity) 329
(MH+, 63), 185 (100), 137 (28), 93 (84); exact mass (FAB+) m/z calcd
for C23H25N2 (MH+) 329.2018, found 329.2018; de >98 (1H NMR),
ee >98% (by comparing the rotation value with a sample obtained
by method B). Anal. Calcd for C23H24N2 (328.2): C, 84.11; H, 7.37; N,
8.53. Found: C, 83.53; H, 7.06, N, 8.38.
4.7.6. (+)-2-[(2R,5R)-2,5-Diphenylpyrrolidin-1-ylmethyl]-6-meth-
ylpyridine 38
NMR (500 MHz, CDCl3)
d
8.61 (d, J = 5.5 Hz, 1H), 8.00 (d,
(Method B) 2-Hydroxymethyl-6-methylpyridine (0.55 g,
4.5 mmol) was allowed to react with PBr3 (1.21 g, 4.5 mmol) in
refluxing CH2Cl2 (10 mL) for 2 h. After TLC (SiO2, hexanes/
ether = 1:1) showed no indication of the remaining starting mate-
rial, the mixture was cooled to rt and neutralized with 30% aque-
ous ammonia. The product was extracted into CH2Cl2
(2 ꢄ 40 mL), washed with brine (2 ꢄ 25 mL), dried (MgSO4), and
filtered to afford a clear solution. Acetonitrile (40 mL) was added
and the solvent was partly removed in vacuo until the total volume
was about 40 mL (most CH2Cl2 was removed). 12 (0.50 g,
2.2 mmol) and K2CO3 (0.62 g, 4.4 mmol) were added to the solu-
tion and the mixture was stirred at 40 °C for 24 h. After aqueous
workup the crude product was purified by flash column chroma-
tography (SiO2, hexane/ether = 1:1) to afford 38 as a pale yellow
J = 2.5 Hz, 1H), 7.70 (dd, J = 2.5, 5.5 Hz, 1H), 7.25 (m, 10H), 4.37
(m, 2H), 3.70 (d, J = 16.5 Hz, 1H), 3.63 (d, J = 16.5 Hz, 1H), 2.60
(m, 2H), 2.09 (m, 2H); 13C NMR (62.9 MHz, CDCl3) d 164.6, 150.7,
143.0, 128.4, 127.9, 127.3, 126.3, 115.0, 113.7, 66.5, 53.8, 33.5;
½
a 2D3
ꢃ
¼ þ61:5 (c 1.3, CH2Cl2); MS (FAB+) m/z (rel intensity) 360.3
(MH+, 6), 246.2 (9). 185.1 (100), 93.2 (77); exact mass (FAB+) m/z
calcd for C22H22N3O2 (MH+) 360.1712, found 360.1703; de >98%
(1H NMR), ee >98% (since there is no sign of degradation of stereo-
chemistry indicated by the high de, the ee was determined to be
higher than 98% depending on the optical purity of 12 which was
measured by the literature method33).
4.7.4. (+)-2-[(2R,5R)-2,5-Diphenylpyrrolidin-1-yl]methyl-4-meth-
oxypyridine 37
solid (0.59 g, 80%). mp = 57–58 °C; ½a D23
¼ þ132 (c 1.15, CHCl3);
ꢃ
(Method B) 2-Hydroxymethyl-4-methoxypyridine (0.36 g,
2.6 mmol) was allowed to react with PBr3 (0.40 mL, 4.2 mmol) in
refluxing CH2Cl2 (5 mL) for 2 h and cooled to rt. After neutraliza-
tion with 30% aqueous ammonia, the product was extracted into
CH2Cl2 (2 ꢄ 30 mL), dried (MgSO4), and filtered to yield a clear
solution. Both TLC analysis (SiO2, hexanes/ether = 1:1) and 1H
NMR showed no indication of the remaining starting material. Ace-
tonitrile (30 mL) was added and the solvent was partly removed in
vacuo until the total volume was about 30 mL (most CH2Cl2 was
removed). 12 (0.32 g, 1.4 mmol) and K2CO3 (0.60 g, 4.3 mmol) were
added to the solution and the mixture was stirred at 40 °C for 24 h.
After normal aqueous workup, flash chromatography (SiO2, CH2Cl2/
EtOAc = 5:1) of the crude product afforded 37 as an orange oil
(0.35 g, 72%): 1H NMR (500 MHz, CDCl3) d 8.22 (d, J = 5.5 Hz, 1H),
de >98% (1H NMR), ee >98% (since there is no sign of degradation
of stereochemistry indicated by the high de, the ee was determined
to be higher than 98% depending on the optical purity of 12 which
was measured by the literature method33).
4.7.7. (+)-2-[(2R,5R)-2,5-Diphenylpyrrolidin-1-ylmethyl]-4-meth-
oxy-6-methylpyridine 39
(Method B) 2-Hydroxymethyl-4-methoxy-6-methylpyridine
(0.40 g, 2.6 mmol) was allowed to react with PBr3 (0.49 mL,
2.6 mmol) in refluxing CH2Cl2 (10 mL) for 1 h. After TLC (SiO2, hex-
ane/ether = 1:1) showed no indication of remaining starting mate-
rial, the mixture was cooled to rt and neutralized with 30%
aqueous ammonia. The product was extracted into CH2Cl2
(2 ꢄ 30 mL), washed with brine (2 ꢄ 15 mL), dried (MgSO4), and

H. Chen et al. / Tetrahedron: Asymmetry 20 (2009) 1672–1682
1681
filtered to yield a clear solution. Acetonitrile (30 mL) was added and
the solvent was partly removed in vacuo until the total volume was
about 30 mL (most CH2Cl2 was removed). 12 (0.46 g, 2.1 mmol) and
K2CO3 (0.62 g, 4.4 mmol) were added to the solution and the mix-
ture was stirred at 40 °C for 24 h. After normal aqueous workup,
flash chromatography (SiO2, hexanes/ether = 1:1) of the crude
product afforded recovered 12 (0.16 g) as a pale yellow solid and
39 as a pale yellow oil (0.43 g, 92% based on consumed 12): 1H
NMR (250 MHz, CDCl3) d 7.22 (m, 10H), 6.90 (d, J = 2.2 Hz, 1H),
6.43 (d, J = 2.2 Hz, 1H), 4.36 (m, 2H), 3.82 (s, 3H), 3.57 (d,
J = 16.1 Hz, 1H), 3.37 (d, J = 16.1 Hz, 1H), 2.55 (m, 2H), 2.34 (s,
3H), 2.00 (m, 2H); 13C NMR (75 MHz, CDCl3) d 166.5, 161.7, 158.6,
143.5, 128.2, 127.9, 126.9, 106.4, 105.3, 65.8, 54.9, 53.2, 33.2,
4.7.9. (+)-2-[(2R,5R)-2,5-Bis(4-methylphenyl)pyrrolidin-1-yl-
methyl]pyridine 41
(Method A) A solution of (1S,4S)-1,4-bis(4-methylphenyl)-1,4-
butandiol 4 (0.53 g, 2.0 mmol), triethylamine (1.05 mL, 10 mmol),
and freshly distilled CH2Cl2 (20 mL) was slowly added to a cold
(ꢀ40 °C) solution of methanesulfonyl chloride (0.55 mL, 4.8 mmol)
in freshly distilled CH2Cl2 (20 mL). The mixture was stirred at
ꢀ20 °C under N2 for 30 min then quenched with saturated NH4Cl
solution (5 mL). The mixture was allowed to warm to rt, and the
solvent was removed in vacuo until the volume was about 5 mL.
The solution was diluted with ethyl acetate (90 mL), and washed
with 1:2:1 water/brine/satd NaHCO3 (4 ꢄ 15 mL), satd NaHCO3
(2 ꢄ 15 mL), and dried (MgSO4). After filtration, the solvent was re-
moved in vacuo until the volume was about 5 mL. The solution was
then cooled to 0 °C, and hexane (20 mL) was added to the solution
dropwise with stirring to form a white precipitate. After the sol-
vent was decanted, fresh hexane (5 mL) was added and the system
was cooled to 0 °C and kept in the dark. 2-Aminomethylpyridine
(2.5 mL, 23 mmol) was added and the resulting mixture was stir-
red under N2 at 0 °C for 2 h and then at rt for 2 d in the dark. The
mixture was extracted with hexane (5 ꢄ 15 mL) and the combined
hexane layers were washed with satd NaHCO3 (2 ꢄ 10 mL), brine
(2 ꢄ 10 mL), and dried (MgSO4). Removal of the solvent after filtra-
tion gave a yellow oil. Flash chromatography (SiO2, hexane/
ether = 1:1) of the oil gave a pale yellow oil which solidified in
the freezer to afford a colorless solid (0.18 g, 26.4%,%de = 75.5).
Recrystallization of this product from hexane afforded 41 as color-
less crystals (83 mg, 12%): 1H NMR (400 MHz, CDCl3) d 8.41 (ddd,
J = 0.9, 1.8, 4.9 Hz, 1H), 7.60 (ddd, J = 1.8, 7.7, 7.8 Hz, 1H), 7.43 (d,
J = 7.8 Hz, 1H) 7.09 (m, 9H), 4.30 (t, J = 4.2 Hz, 2H), 3.62 (d,
J = 15.7 Hz, 1H), 3.32 (d, J = 15.7 Hz, 1H), 2.55 (m, 2H), 2.31 (s,
6H), 2.01 (m, 2H); 13C NMR (100 MHz, CDCl3) d 149.5, 137.6,
136.9, 129.8, 128.9, 126.8, 123.5, 122.2, 66.4, 54.0, 34.0, 21.9;
24.5; ½a 2D3
ꢃ
¼ þ94:7 (c 1.28, CH2Cl2); MS (FAB+) m/z (rel intensity)
359 (MH+, 48), 185 (100), 137 (30), 93 (73); exact mass (FAB+) m/
z calcd for C24H27N2O (MH+) 359.2123, found 359.2118; de >98
(1H NMR), ee >98% (since there is no sign of degradation of stereo-
chemistry indicated by the high de, the ee was determined to be
higher than 98% depending on the enantiomeric purity of 12 which
was measured by a literature method33). Anal. Calcd for C24H26N2O
(358.5): C, 80.41; H, 7.31; N, 7.81. Found: C, 80.21; H, 7.28; N, 7.78.
4.7.8. (+)-2-[(2R,5R)-(2,5-Diphenylpyrrolidin-1-ylmethyl)]quino-
line 40
(Method B) 2-Quinolinecarboxylic acid (29, 1.04 g, 6.0 mmol)
and thionyl chloride (10 mL, 137 mmol) in a round-bottomed flask
was allowed to reflux under N2 for 2 h. After the mixture was
cooled to rt, excess thionyl chloride was removed in vacuo to afford
an orange solid. This solid was cooled to 0 °C, and dry methanol
(15 mL, excess) was added dropwise with stirring. The resulting
solution was heated and allowed to reflux for 3 h, after which time
the solvent was removed in vacuo to afford a yellow solid. After the
flask containing the solid was cooled to 0 °C, 95% ethanol (20 mL)
was added and NaBH4 (0.91 g, 24 mmol) was added in portions.
The mixture was stirred under N2 overnight until TLC (SiO2, hex-
ane/ether = 1:1) showed no indication of the starting material.
The solid was then filtered and solvent in the filtrate was removed
in vacuo to afford a yellow oil. Flash chromatography (SiO2, ether)
of this oil afforded 2-hydroxymethylquinoline 30 as a yellow oil
(0.50 g, 51%) with1H NMR data matching that reported.56 Commer-
cially available HBr (48%, 10 mL) was then added to this compound
with stirring at 0 °C and then the mixture was allowed to reflux un-
der N2 for 3 h. After the mixture was cooled to rt, it was neutralized
to pH 6–8 with 30% ammonium hydroxide. The resulting aqueous
solution was extracted with CH2Cl2 (2 ꢄ 10 mL), dried (MgSO4),
and concentrated to afford 2-bromomethylquinoline 3147 as a
brown oil (80% from 30. 12 (0.49 g, 2.2 mmol), CH3CN (25 mL)
and K2CO3 (0.53 g, 3.8 mmol) were added and the mixture was stir-
red at 40 °C under N2 for 2 d. After normal workup, the crude prod-
uct was purified by flash chromatography (SiO2, CH2Cl2/
EtOAc = 40:1) to give recovered 12 (0.25 g) as a yellow solid and
40 as a pale yellow solid (0.56 g, 75% based on consumed 12): 1H
NMR (500 MHz, CDCl3) d 8.41 (ddd, J = 0.9, 1.8, 4.9 Hz, 1H), 7.60
(ddd, J = 1.8, 7.7, 7.8 Hz, 1H), 7.43 (d, J = 7.8 Hz, 1H) 7.09 (m, 9H),
4.30 (t, J = 4.2 Hz, 2H), 3.62 (d, J = 15.7 Hz, 1H), 3.32 (d,
J = 15.7 Hz, 1H), 2.55 (m, 2H), 2.31 (s, 6H), 2.01 (m, 2H); 13C NMR
(125 MHz, CDCl3) d 149.5, 137.6, 136.9, 129.8, 128.9, 126.8,
mp = 145–146 °C; ½a D23
ꢃ
¼ þ159:7 (c 0.96, CHCl3); MS (FAB+) m/z
(rel intensity) 343 (MH+, 75), 185 (100), 137 (28), 93 (90); exact
mass (FAB+) m/z calcd for C24H27N2 (MH+) 343.2174, found
343.2174; de >98% (1H NMR), ee >99% [Chiralcel OD, hexanes/iso-
propanol/triethylamine = 99:1:0.1, 0.5 mL/min, 18.2 min (S,S),
18.6 min (meso), 19.0 min (R,R)]. Anal. Calcd for C24H26N2 (343.2):
C, 84.17; H, 7.65; N, 8.18. Found: C, 83.23; H, 7.35, N, 8.22.
4.7.10. (+)-2-[2,5-Bis-(4-tert-butylphenyl)pyrrolidin-1-ylmethyl]-
pyridine 42
(Method B) 2-Hydroxymethylpyridine (0.16 g, 1.5 mmol) was
allowed to react with PBr3 (0.28 mL, 3.0 mmol) in refluxing CH2Cl2
(20 mL) for 1 h. After TLC (SiO2, hexanes/ether = 1:1) showed no
indication of remaining starting material, the mixture was cooled
to rt and neutralized with 30% aqueous ammonia. The product
was extracted into ether (2 ꢄ 50 mL), washed with brine
(2 ꢄ 15 mL), dried (MgSO4), and filtered to yield a clear solution.
Acetonitrile (30 mL) was added and the solvent was partly re-
moved in vacuo until the total volume was about 30 mL. Com-
pound 15 (0.39 g, 1.2 mmol) and K2CO3 (0.62 g, 4.4 mmol) were
added to the solution and the mixture was stirred at 40 °C for
24 h. After normal aqueous workup, flash chromatography (SiO2,
hexanes/ether = 2:1) of the crude product afforded 42 as an off-
white solid (0.27 g, 55%). 1H NMR (250 MHz, CDCl3) d 8.39 (ddd,
J = 1.7, 1.7, 5.8 Hz, 1 H), 7.51 (ddd, J = 1.7, 7.6, 7.6 Hz, 1 H), 7.41
123.5, 122.2, 66.4, 54.0, 34.0, 21.9; ½a D28
¼ þ141:2 (c 1.8, CHCl3);
ꢃ
MS (FAB+) m/z (rel intensity) 365 (MH+, 100), 246 (13), 185 (88),
93 (54), 75 (8); exact mass (FAB+) m/z calcd for C26H25N2 (MH+)
365.2018, found 365.2019; de >98% (1H NMR), ee >98% (since there
is no sign of degradation of stereochemistry indicated by the high
de, the ee was determined to be higher than 98% depending on the
enantiomeric purity of 12 which was measured by a literature
method33). Anal. Calcd for C26H24N2 (364.48): C, 85.68; H, 6.64;
N, 7.69. Found: C, 85.28; H, 6.28; N, 7.58.
(dd, J = 1.7, 7.6 Hz, 1 H), 7.34 (d, J = 8.3 Hz, 4 H), 7.16 (d,
J = 8.3 Hz, 4 H), 6.98 (dd, J = 5.8, 7.6 Hz, 1 H), 4.31 (t, J = 4.8 Hz, 2
H), 3.63 (d, J = 15.8 Hz, 1 H), 3.40 (d, J = 15.8 Hz, 1 H), 2.52 (m, 2
H), 2.02 (m, 2 H), 1.28 (s, 18 H); 13C NMR (62.9 MHz, CDCl3) d
160.6, 149.4, 148.5, 139.9, 135.8, 127.6, 124.9, 122.2, 121.0, 64.9,
53.2, 34.3, 33.0, 31.3; mp = 64–65 °C; ½a D23
¼ þ125:4 (c 1.15,
ꢃ
CH2Cl2); MS (FAB+) m/z (rel intensity) 427 (MH+, 40), 338 (4), 246
(12), 185 (100), 93 (58); exact mass (FAB+) m/z calcd for

1682
H. Chen et al. / Tetrahedron: Asymmetry 20 (2009) 1672–1682
C30H39N2 (MH+) 427.3113, found 427.3110; de >98 (1H NMR), ee
>98% (since there is no sign of degradation of stereochemistry indi-
cated by the high de, the ee was determined to be higher than 98%
depending on the enantiomeric purity of 15). Anal. Calcd for
C30H38N2 (426.3): C, 84.46; H, 8.98; N, 6.57. Found: C, 84.31; H,
9.00; N, 6.55.
21. Cahill, J. P.; Bohnen, F. M.; Goddard, R.; Krueger, C.; Guiry, P. J. Tetrahedron:
Asymmetry 1998, 9, 3831–3839.
22. Cahill, J. P.; Cunneen, D.; Guiry, P. J. Tetrahedron: Asymmetry 1999, 10, 4157–
4173.
23. Cahill, J. P.; Lightfoot, A. P.; Goddard, R.; Rust, J.; Guiry, P. J. Tetrahedron:
Asymmetry 1998, 9, 4307–4312.
24. Cahill, J. P.; Guiry, P. J. Tetrahedron: Asymmetry 1998, 9, 4301–4305.
25. Farrell, A.; Goddard, R.; Guiry, P. J. J. Org. Chem. 2002, 67, 4209.
26. Drommi, D.; Saporita, M.; Bruno, G.; Faraone, F.; Scafato, P.; Rosini, C. Dalton
Trans. 2007, 1509.
27. Chelucci, G.; Pinna, G. A.; Saba, A. Tetrahedron: Asymmetry 1997, 8, 2571–2578.
28. Fiore, K.; Martelli, G.; Monari, M.; Savoia, D. Tetrahedron: Asymmetry 1999, 10,
4803.
29. Ferioli, F.; Fiorelli, C.; Martelli, G.; Monari, M.; Savoia, D.; Tobaldin, P. Eur. J. Org.
Chem. 2005, 1416–1426.
30. Kwong, H.-L.; Yeung, H.-L.; Yeung, C.-T.; Lee, W.-S.; Lee, C.-S.; Wong, W.-L.
Coord. Chem. Rev. 2007, 251, 2188.
31. Kizirian, J.-C. Chem. Rev. 2008, 108, 140–205.
4.8. Typical procedure for palladium-catalyzed allylic alkylations
Palladium-catalyzed allylic alkylations were run according to
the method provided by Allen et al.57,58 The percentage ee of the
addition product was determined by 1H NMR using Eu(tfc)3 as
the chiral shift agent, or by chiral HPLC using a Daicel Chiralcel
OD column.
32. Short, R. P.; Kennedy, R. M.; Masamune, S. J. Org. Chem. 1989, 54, 1755–1756.
33. Chong, J. M.; Clarke, I. S.; Koch, I.; Olbach, P. C.; Taylor, N. J. Tetrahedron:
Asymmetry 1995, 6, 409–418.
Acknowledgments
34. Lieser, J. K. Synth. Commun. 1983, 13a, 765.
35. Otten, S.; Frohlich, R.; Haufe, G. Tetrahedron: Asymmetry 1998, 9, 189–191.
36. Bailey, P. S.; Lutz, R. E. J. Am. Chem. Soc. 1948, 70, 2412–2414.
37. Conant, J. B.; Lutz, R. E. J. Am. Chem. Soc. 1923, 45, 1303–1307.
38. Lutz, R. E. Org. Synth. Coll. 1955, 3, 248–251.
39. Lutz, R. E. J. Am. Chem. Soc. 1930, 52, 3423.
40. Crowell, T. I.; Helsley, G. C.; Lutz, R. E.; Scott, W. L. J. Am. Chem. Soc. 1963, 85,
443–446.
The authors thank the University of Connecticut Research Foun-
dation, the Department of Chemistry, University of Arizona, and
the Graduate Research Final Fund at the University of Arizona for
partial financial support for this research. We also thank Per-Ola
Norrby for helpful discussions.
41. Brown, H. C.; Chandrasekharan, J.; Ramachandran, P. V. J. Am. Chem. Soc. 1988,
110, 1539–1546.
42. Guzikowski, A. P.; Tamiz, A. P.; Acosta-Burruel, M.; Hong-Bae, S.; Cai, S. X.;
Hawkinson, J. E.; Keana, J. F. W.; Kesten, S. R.; Shipp, C. T.; Tran, M.;
Whittemore, E. R.; Woodward, R. M.; Wright, J. L.; Zhou, Z.-L. J. Med. Chem.
2000, 43, 984–994.
43. Ashimori, A.; Ono, T.; Uchida, T.; Ohtaki, Y.; Fukaya, C.; Watanabe, M.;
Yokoyama, K. Chem. Pharm. Bull. 1990, 38, 2446–2458.
44. Tanga, M. J.; Bupp, J. E.; Tochimoto, T. K. J. Heterocycl. Chem. 1997, 34, 717–728.
45. Kuehler, T.; Fryklund, J.; Bergman, N.; Weilitz, J.; Lee, A.; Larsson, H. J. Med.
Chem. 1995, 38, 4906–4916.
46. Szafran, M.; Brycki, B.; Dega-Szafran, Z.; Nowak-Wydra, B. J. Chem. Soc., Perkin
Trans. 2 1991, 1161–1166.
47. Chiu, Y.-H.; dos Santos, O.; Canary, J. W. Tetrahedron 1999, 55, 12069–12078.
48. Kubota, H.; Nakajima, M.; Koga, K. Tetrahedron Lett. 1993, 34, 8135–8138.
49. Crabtree, R. H. The Organometallic Chemistry of the Transition Metals, 2nd ed.;
Wiley Interscience: New York, 1994.
50. Peña-Cabrera, E.; Norrby, P.-O.; Sjögren, M.; Vitagliano, A.; De Felice, V.; Oslob,
J.; Ishii, S.; O’Neill, D.; Åkermark, B.; Helquist, P. J. Am. Chem. Soc. 1996, 118,
4299–4313.
51. Hagelin, H.; Åkermark, B.; Norrby, P.-O. Organometallics 1999, 18, 2884–2895.
52. Hagelin, H.; Åkermark, B.; Norrby, P.-O. Chem. Eur. J. 1999, 5, 902–909.
53. Pfaltz, A. Acc. Chem. Res. 1993, 26, 339–345.
References
1. Whitesell, J. K.; Felman, S. W. J. Org. Chem. 1977, 42, 1663–1664.
2. Pearson, A. J.; Gontcharov, A. V.; Zhu, P. Y. Tetrahedron 1997, 53, 3849–3862.
3. Yamazaki, T.; Welch, J. T.; Plummer, J. S.; Gimi, R. Tetrahedron Lett. 1991, 32,
4267–4270.
4. Genicot, C.; Ghosez, L. Tetrahedron Lett. 1992, 33, 7357–7360.
5. Zhang, Q.; Mohan, R. M.; Cook, L.; Kazanis, S.; Peisach, D.; Foxman, B. M.;
Snider, B. B. J. Org. Chem. 1993, 58, 7640–7651.
6. He, S.; Kozmin, S. A.; Rawal, V. H. J. Am. Chem. Soc. 2000, 122, 190–191.
7. Kozmin, S. A.; Rawal, V. H. J. Am. Chem. Soc. 1997, 119, 7165–7166.
8. Kozmin, S. A.; Rawal, V. H. J. Am. Chem. Soc. 1999, 121, 9562–9573.
9. Gouverneur, V.; Ghosez, L. Tetrahedron Lett. 1991, 32, 5349–5352.
10. Katsuki, T.; Yamaguchi, M. Tetrahedron Lett. 1985, 26, 5807–5810.
11. Kawanami, Y.; Ito, Y.; Kitagawa, T.; Taniguchi, Y. Tetrahedron Lett. 1984, 25,
857–860.
12. Taber, D. F.; Gorski, G. J.; Liable-Sands, L. M.; Rheingold, A. L. Tetrahedron Lett.
1997, 38, 6317–6318.
13. Kim, B. H.; Lee, H. B.; Hwang, J. K.; Kim, Y. G. Tetrahedron: Asymmetry 2005, 16,
1215.
14. Sprott, K. T.; Corey, E. J. Org. Lett. 2003, 5, 2465–2467.
15. Bertilsson, S. K.; Soedergren, M. J.; Andersson, P. G. J. Org. Chem. 2002, 67, 1567–
1573.
16. Sweet, J. A.; Cavallari, J. M.; Price, W. A.; Ziller, J. Z.; McGrath, D. V. Tetrahedron:
Asymmetry 1997, 8, 207–211.
54. Brown, J. M.; Hulmes, D. I.; Guiry, P. J. Tetrahedron 1994, 50, 4493–4506.
55. Mino, T.; Shiotsuki, M.; Yamamoto, N.; Suenaga, T. J. Org. Chem. 2001, 66, 1795–
1797.
56. Rieger, B.; Abu-Surrah, A. S.; Fawzi, R.; Steiman, M. J. Organomet. Chem. 1995,
497, 73–79.
57. Allen, J. V.; Bower, J. F.; Williams, J. M. J. Tetrahedron: Asymmetry 1994, 5, 1895–
1898.
58. Allen, J. V.; Coote, S. J.; Dawson, G. J.; Frost, C. G.; Martin, C. J.; Williams, J. M. J. J.
Chem. Soc., Perkin Trans. 1 1994, 2065–2072.
17. Waernmark, K.; Stranne, R.; Cernerud, M.; Terrien, I.; Rahm, F.; Nordstroem, K.;
Moberg, C. Acta Chem. Scand. 1998, 52, 961–963.
18. Melchiorre, P.; Joergensen, K. A. J. Org. Chem. 2003, 68, 4151–4157.
19. Hoen, R.; Berg, M. v. d.; Bernsmann, H.; Minnaard, A. J.; Vries, J. G. D.; Feringa, B.
L. Org. Lett. 2004, 6, 1433–1436.
20. Stranne, R.; Moberg, C. Eur. J. Org. Chem. 2001, 2001, 2191–2195.