An experimental and theoretical study on free ligand conformational preferences and enantioselectivity relationship for the asymmetric addition of diethylzinc to benzaldehyde

Article abstract of DOI:10.1016/j.tetasy.2010.03.009

An experimental and theoretical study on free ligand conformational preferences and enantioselectivity relationship has been described for the asymmetric addition of diethylzinc to benzaldehyde. The results show that a correlation must exist between the ground-state ligand conformational populations and the observed ee values in this reaction. As the populations of the free ligand conformation (the desired conformation) in favor of the improvement of the reaction enantioselectivity increase, so does the reaction enantioselectivity. However, the desired conformation must not be the preferred one of the ground-state ligand. This conformation-enantioselectivity relationship is well explained based on a zinc amino-alkoxide (a true asymmetric catalyst). The final synthesis and assessment of the new chiral catalyst in the asymmetric addition of Et₂Zn to benzaldehyde revealed that this necessary relationship guided our design of highly enantioselective ligands or rational improvement of existing ligands by means of knowledge of conformational analysis.

Full text of DOI:10.1016/j.tetasy.2010.03.009

Tetrahedron: Asymmetry 21 (2010) 486–493
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
An experimental and theoretical study on free ligand conformational
preferences and enantioselectivity relationship for the asymmetric
addition of diethylzinc to benzaldehyde
*
Min-Can Wang , Zhi-kang Liu, Song Li, Xue Ding, Yuan Li, Ming-Sheng Tang
Department of Chemistry, Zhengzhou University, Zhengzhou, Henan 450 052, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
An experimental and theoretical study on free ligand conformational preferences and enantioselectivity
relationship has been described for the asymmetric addition of diethylzinc to benzaldehyde. The results
show that a correlation must exist between the ground-state ligand conformational populations and the
observed ee values in this reaction. As the populations of the free ligand conformation (the desired con-
formation) in favor of the improvement of the reaction enantioselectivity increase, so does the reaction
enantioselectivity. However, the desired conformation must not be the preferred one of the ground-state
ligand. This conformation-enantioselectivity relationship is well explained based on a zinc amino-alkox-
ide (a true asymmetric catalyst). The final synthesis and assessment of the new chiral catalyst in the
asymmetric addition of Et2Zn to benzaldehyde revealed that this necessary relationship guided our
design of highly enantioselective ligands or rational improvement of existing ligands by means of knowl-
edge of conformational analysis.
Received 26 January 2010
Accepted 9 March 2010
Available online 10 April 2010
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
the efficiency of the research process,2 however, they do not pro-
vide with regard to the origin of enantioselectivities. This knowl-
Over the last three decades, many of the most general and effi-
cient methods for the synthesis of enantiomerically enriched com-
pounds have depended on chiral metal catalysts. Much effort has
been devoted to the design of appropriate chiral ligands capable of
efficient chirality transfer,1 since the chiral ligands are responsible
not only for the activation of the metal atom where the catalytic
activity of the reaction resides but also for the generation of a chiral
environment around the metal atom, which controls the enantiose-
lectivity of the reaction. One important question is how one goes
about identifying an optimal chial ligand? The most common meth-
od for discovering an efficient chiral catalyst seems to be one of the
trial and error, wherein, a reasonable lead catalyst is first found.
Then, based on intuition or chemical knowledge about mechanistic
studies of the reaction being catalyzed, numerous catalyst struc-
tures are prepared by placing different substituents on the substruc-
ture of the initial lead catalyst. Finally, catalyst candidates are
screened in terms of the evaluation of activity and selectivity, and
an optimal catalyst is identified after months or years. Therefore,
the process of finding an effective chiral catalyst has occupied the
time of many investigators over the last three decades.
edge, however, is essential for the rational design of a chiral
catalyst.
Computational chemistry methods have been extensively used
to rationalize the origin of the observed enantioselectivity based
on transition structure calculations.3 However, only a very few
examples have been reported that have attempted to predict the
stereochemical results of reactions,4 with even fewer uses with re-
gard to the design of new chiral catalysts.5 The main reason for the
lack of progress may be (1) a detailed knowledge of the reaction
mechanism is required; (2) accurate parameters are lacking for
metal complexes, which are necessary to model metal-catalyzed
reactions; and (3) many variables must be assessed, which requires
extensive time, effort, and resources.
Therefore, it is highly desirable to develop simple, practical, and
efficient approaches for finding a highly enantioselective chiral cat-
alyst. Herein, we initially describe our observations: the necessary
relationship which exists between the free ligand conformational
populations and enantioselectivity in the asymmetric addition of
diethylzinc to benzaldehyde. Subsequent theoretical computation
supports our observation. The final synthesis and evaluation of a
new chiral ligand in the asymmetric addition of organozinc to alde-
hydes reveal further that this necessary relationship can guide our
design of highly enantioselective ligands, or rational improvement
of existing ligands with only the knowledge of conformational
analysis.
Combinatorial approaches in association with high throughput
screening have already been put forth to accelerate and increase
* Corresponding author. Tel./fax: +86 371 67767895.
E-mail address: wangmincan@zzu.edu.cn (M.-C. Wang).
0957-4166/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2010.03.009

M.-C. Wang et al. / Tetrahedron: Asymmetry 21 (2010) 486–493
487
We have selected this catalytic system to study because a con-
siderable number of catalysts have been designed and evaluated
for their ability to induce asymmetry,6 but more importantly be-
cause a relationship between the free ligand conformational popu-
lations and enantioselectivity has already been noted by Zhu et al.7
In their work, due to the flexibility of the backbone of a chiral li-
gand, numerous conformers, resulting in a flip of a five- or six-
membered ring, pyramidal inversion at nitrogen, and rotation
about the exocyclic bonds might exist at ambient room tempera-
ture. All the likely minimum energy conformations (72 total possi-
bilities) at the HF/3-21G level were initially examined in order to
find the most stable cis- and trans-conformations. As a result of
these complexities, what they discovered was that the observed
ee values, in most cases, were only related to the populations of
the cis- versus trans-free ligand conformations formed by pyrami-
dal inversion at nitrogen, which led to a reversal of the absolute
configuration of the stereogenic nitrogen. To the best of our knowl-
edge, there have been no reports on the effect of ground-state li-
gand conformational equilibria on the observed enantioselectivity.
enantioselectivity from 92.6% to 96.0% ee. On the basis of our anal-
ysis of the reaction pathway and the proposed possible transition
states,8c the methyl group situated far away from the reaction ac-
tive sites did not seem to have an impact on the reaction enanti-
oselectivity. Why does the introduction of the methyl substituent
result in an increase in enantioselectivity? Further insight into
these chiral ligands revealed that the introduction of a methyl
group on the aziridine ring resulted in the differences in relative
free conformational populations of 1 and 2, that is, the introduction
of a methyl group did not affect the flexibility of the framework,
yet it led to the equilibrium redistributions of the conformational
isomers. This observation suggested that a relationship must exist
between the free ligand conformational populations and the ob-
served ee values in the asymmetric addition of diethylzinc to
benzaldehyde.
For the sake of better understanding, the structural features of
aziridino amino alcohol compounds are first described as follows.
In (2S)-1-(ferrocenylmethyl)aziridin-2-yl(diphenyl)methanol
1
(Fig. 1),8a the aziridine ring has completely planar-ring structure,
and thereby free of ring flip conformation. Due to the formation
of the intramolecular hydrogen bond (in a five-membered ring
structure, Fig. 1) as well as the strong repulsive interaction be-
tween the bulky ferrocenyl group on the nitrogen atom of the azi-
ridine ring and the bulky diphenylhydroxymethyl group on the
three-membered ring, nitrogen pyramidal inversion is effectively
blocked. As a result, the nitrogen atom on the aziridine ring also
becomes a stereocenter. The crystal structure of compound 1
revealed that the value of the torsion angle N(1)–C(13)–C(14)–
C(15) in 1 is 90.0(3)° (Fig. 2),8a suggesting that one phenyl substi-
2. Results and discussion
In recent years,8 we have been exploring the use of chiral ferro-
cene-based small-ring heterocycle ligands containing a b-amino
alcohol moiety in the catalytic asymmetric addition of organozinc
to aldehydes. In our previous work,8a,c we reported the synthesis of
chiral ferrocenyl aziridino alcohol ligands 1 and 2 (Fig. 1) and their
applications in the enantioselective addition of diethylzinc to
benzaldehyde. The introduction of a methyl group on aziridine
ring, with the C-3 methyl group anti to the C-2 bulky diph-
enylhydroxymethyl group, led to a further improvement in the
tuent on the a-carbon occupies an axial position with regard to the
five-membered ring, which we called the axial phenyl group. The
values of the torsion angles N(1)–C(13)–C(14)–C(21) and C(12)–
C(13)–C(14)–C(21) in 1 are ꢀ147.6(3)° and ꢀ78.7(4)°, respectively,
indicating that the other phenyl group is located in the equatorial
position, which we called the equatorial phenyl group, where this
arrangement minimizes the steric interaction between the equato-
rial phenyl group and the aziridine ring moiety. As a result of the
aforementioned reasons a fixed conformation about the bulky
diphenylhydroxymethyl group results in the aziridine ring at the
C-2 position. Such structural features are common for those aziri-
dino amino alcohol derivatives which bear two bulky phenyl
Ph
Ph
Ph
Ph
O
O
H
3
H
3
H
N
H
N
2
2
H
H
H
Me
1
1
H
H
Fc
Fc
H
H
1
2
Figure 1. The structures of chiral ligands 1 and 2.
Figure 2. Perspective view of compound 1. Selected bond distances (Å), angles (°), and torsion angles (°) are N(1)–C(13), 1.458(4); N(1)–C(12), 1.460(4); C(12)–C(13),
1.454(4); C(13)–C(14), 1.537(5); O(1)–C(14), 1.423(4); O(1)–H(1), 0.89(5); O(1)–N(1), 2.678(5); N(1)–H(1), 1.977(5). C(13)–N(1)–C(12), 59.8(2); C(13)–C(12)–N(1), 60.1(2);
C(12)–C(13)–N(1), 60.1(2); N(1)–C(13)–C(14), 114.5(3); C(14)–O(1)–H(1), 97(3); N(1)–H(1)–O(1), 134. N(1)–C(13)–C(14)–C(15), 90.0(3); N(1)–C(13)–C(14)–C(21),
ꢀ147.6(3); C(12)–C(13)–C(14)–C(21), ꢀ78.7(4); C(13)–N(1)–C(11)–C(10), ꢀ147.4(3); C(12)–N(1)–C(11)–C(10), ꢀ80.6(4); N(1)–C(13)–C(14)–O(1), ꢀ30.1(4).

488
M.-C. Wang et al. / Tetrahedron: Asymmetry 21 (2010) 486–493
groups at the
a
-position.8a,9 The value of the torsion angle C(13)–
replaced by a methyl group to yield derivative 2, there are three
possible conformations 2a–2c (Fig. 4). Conformation 2b is the most
stable due to minimizing the steric interaction between the bulky
ferrocenyl group and the methyl group. Therefore, the conforma-
tional isomer 2b is the predominant one in the equilibrium
mixture.
To further examine the equilibrium distributions of the confor-
mational isomers, the relative energies of the conformers of 1 and
2 were calculated using GAUSSIAN 03 program.10 All geometries were
optimized at the HF level using the 6-31G basis set for C, H, N, and
O, LanL2DZ basis set for Fe. The single-point energy was deter-
mined at B3LYP/6-31+g (d,p). The calculated results are summa-
rized in Table 1 (entries 1–6). As seen from Table 1, the order of
N(1)–C(11)–C(10) in 1 is ꢀ147.4(3)°, suggesting that a nonbonded
interaction exists between the bulky ferrocenyl group and the axial
phenyl substituent on
cal shape.
a-carbon because of the ferrocenyl cylindri-
Due to the aforementioned reasons, the chiral compound 1 has
three possible conformations about the exocyclic N–C bond-free
rotation (1a–c, Fig. 3). In conformation 1c, the ferrocenyl group lies
under the aziridine ring, while in conformations 1a and 1b, a
hydrogen atom lies under the aziridine ring. Because of the non-
bonded repulsion between the bulky ferrocenyl group and the azi-
ridine ring unit in 1c, and the interaction between the bulky
ferrocenyl group and the axial phenyl substituent on
a-carbon in
1b, conformation 1a is more stable than conformations 1b and
1c. Thus, in the equilibrium mixture, conformer 1a is the predom-
inant one, which is in agreement with the X-ray crystal structure of
the ligand 1.
When a hydrogen atom of the aziridine ring at the 3-position,
which is positioned anti to the diphenylhydroxymethyl group, is
relative energies for conformers of the ligands 1 and 2 is
1a < 1b < 1c (relative conformational populations of 1a/1b/
1c = 50.0:27.5:22.5) and 2b < 2a < 2c (2a/2b/2c = >99.9:0:0),
respectively. The calculated results were in agreement with con-
formational analysis. The reliability of an ab initio approach to
examine the relative stability of conformers had previously been
Ph
Ph
H
Ph
H
H
Ph (equatorial)
O
H
H
N
Ph (axial)
Ph
Ph
Ph
HO
Fc
HO
HO
H
N
H
N
H
H
H
N
Fc
H
H
H
Fc
H
H
Fc
1a
1b
1c
1
Figure 3. Three conformations about the N–C bond rotation for ligand 1.
Ph
Ph
Ph
H
H
H
H
H
Ph
Ph
H
O
H
H
Ph
Ph
Ph
Me
Me
Me
HO
Fc
HO
H
HO
H
H
N
Me
H
N
N
N
H
H
Fc
Fc
H
H
Fc
H
2c
2a
2
2b
Figure 4. Three conformations about the N–C bond rotation for ligand 2.
Table 1
The relationship between the free ligand conformational equilibria and the observed enantioselectivity in the addition of diethylzinc to benzaldehyde
Ph
O
CPh2OH
H
CPh2OH
H
H
R
H
R
CPh2OH
H
H
H
R
H
N
Ph
Fc
N
H
N
H
N
R
H
H
H
H
H
Fc
Fc
H
Fc
H
c
b
a
Ratiob (%)
ee
a
Entry
Ligands
R
Comf
E (Hartree)
Erel (kJ/mol)
1
2
3
4
5
6
1
H
H
H
Me
Me
Me
1a
1b
1c
2a
2b
2c
ꢀ1259.21374
ꢀ1259.21319
ꢀ1259.21296
ꢀ1298.52079
ꢀ1298.53118
ꢀ1298.51769
0.0
1.5
2.0
27.3
0.0
35.4
50.0
27.5
22.5
0.0
>99.9
0.0
92.7
2
96.0
a
Relative to the lowest energy conformation (shown as zero kJ/mol).
Conformational population is obtained from a Boltzmann distribution of relative energies at 298 K.
b

M.-C. Wang et al. / Tetrahedron: Asymmetry 21 (2010) 486–493
489
acknowledged.11 In addition, solvent calculations using the IPCM
model for the conformers qualitatively agree with the gas-phase
ab initio results.11d
A comparison of the enantioselectivity (96.0% ee) induced by 2
with that (92.6% ee) by 1 shows that the introduction of a methyl
group that is positioned anti to the diphenylhydroxymethyl group
leads to an improvement in the enantioselectivity when used as
the catalyst in the addition of diethylzinc to benzaldehyde. The
enantioselective differences afforded by 1 and 2 could be attrib-
uted to the difference of the relative free conformational popula-
on the diphenylhydroxymethyl group point toward the same
direction with respect to the five-membered coordination ring.
As a result, the ferrocenyl and phenyl groups in b, due to the coop-
eratively directing effect of the two substituents, can block more
effectively the approach of benzaldehyde and diethylzinc from this
face, when compared with the hydrogen atom and phenyl substit-
uents in a and c. That is, the conformer b of the chiral catalysts
efficiently distinguishes between the face of the five-membered
Zn-chelate ring, which resulted in the improvement of the enanti-
oselectivity. Therefore, the enantioselectivity will be enhanced
when the amount of the desired conformer b increases.
tions in
1 and 2. Compared with the relative ground-state
conformational populations (1b/1a ratios as 0.55) in 1, the sub-
stantial increase in free conformational populations (>99.9%) of
2b resulted in an improvement in the enantioselectivity, that is,
the observed enantioselectivity increases when increasing the
amount of the conformation isomer 2b. Therefore, the structure
of the conformation 2b, where the relative position of the bulky
ferrocenyl group is oriented toward the same direction of the axial
phenyl group, as in 2b, favored the enhancement of the reaction
enantioselectivity. For the sake of convenience, the conformation
in favor of the relative improvement of the reaction enantioselec-
tivity is called the desired conformation. These analyses and results
suggest that a relationship must exist between the free ligand con-
formational populations and the observed ee values in the asym-
metric addition of diethylzinc to benzaldehyde. In addition, the
preferred conformation of the free ligand must not be the desired
conformation.
This conformation-enantioselectivity relationship is under-
standable. Noyori et al. have studied the mechanism of this reac-
tion extensively, both theoretically and experimentally.12 The
true asymmetric catalyst is believed to be a zinc amino-alkoxide
(a planar Zn structure, Fig. 5). Compared with free ligands, the
replacement of the chelated proton with an ethylzinc moiety
should not lead to any significant structural distortion because li-
gand ground-state conformation resembles that in the complex.
Theoretical computation also demonstrated that stable five-mem-
bered Zn-chelate ring conformations resembled those of the free
ligand for aziridino alcohol analogue.13 Therefore, the chiral cata-
lyst, generated by the reaction of the conformationally flexible
ligands 1 and 2 with diethylzinc, has also three possible conforma-
tions about the corresponding N–C bond-free rotation (a–c, Fig. 5).
In conformation b, the ferrocenyl group and the axial phenyl group
The next question one might ask is does the same conforma-
tion-enantioselectivity relationship exist between the free ligand
conformational populations and the ee values obtained in the
asymmetric addition of diethylzinc to benzaldehyde when the ferr-
ocenylmethyl group on the nitrogen atom of aziridine-based skel-
eton was replaced by other substituents? This question is very
important because if the necessary correlation exists, one can envi-
sion a number of different procedures to design the preferred con-
formation that will increase the reaction enantioselectivity by
intentionally modifying a conformationally flexible ligand’s struc-
ture, which is called the conformational design of flexible
molecules.14
In previous literature,9b,15 chiral ligands 3–6 with the same
backbone as 1 and 2 were reported which were used in the asym-
metric addition of diethylzinc to benzaldehyde with large differ-
ences in ee values. However, these big differences in ee values
have to date not been rationalized. Inspection of such structures
revealed that the observed enantioselectivity also correlated with
the equilibrium distributions of ground-state conformers.
For chiral ligand 3 (Fig. 6),15a which replaces the ferrocenyl
group in 1 with a phenyl substituent, the relative energy order of
conformational isomers for the ligand 3 was 3a < 3b < 3c (relative
populations: 3a/3b/3c = 66.8:21.4:11.8, Table 2, entries 1–3). This
shows that conformer 3a is the preferred one, which is in agree-
ment with the case of the chiral ligand 1. However, the chiral li-
gand 3 only gave 49.0% ee, whereas the chiral ligand 1 afforded
up to 92.7% ee. This big difference (92.7% vs 49.0%) in enantioselec-
tivity is due to the steric effect of the bulky ferrocenyl group com-
pared with the phenyl substituent, even though the ferrocenyl
group in conformer 1a can also exert steric hindrance because of
its cylindrical shape (relative to the planar phenyl group). These re-
sults indicate that the observed enantioselectivity was related to
not only the amount of the desired conformation but also the size
of the substituent steric hindrance.
The replacement of the hydrogen atom of 3 at the 3-position be-
low the aziridine ring with a methyl group afforded chiral ligand 4
(Fig. 7).15a For ligand 4, the energy ordering with the 4a < 4b < 4c
sequence (relative populations: 4a/4b/4c = 72.2:27.4:0.3, Table 2,
entries 4–6) was the same as that of 3. However, the relative con-
formational populations (4b/4a ratios as 0.38) in 4 were higher
than those (3b/3a ratios as 0.32) in 3. The larger populations of the
desired conformation 4b relative to 3b are responsible for the higher
Ph
Ph
Ph
Et
Et
Et
O
O
O
H
R
H
R
H
R
Zn
N
Zn
N
Zn
N
Ph
Ph
Ph
H
H
H
H
Fc
H
Fc
H
H
H
H
Fc
b
c
a
Figure 5. The conformers of the true asymmetric catalyst.
Ph
Ph
Ph
H
H
H
Ph
O
H
H
N
Ph
Ph
Ph
Ph
HO
Ph
HO
HO
H
N
H
N
H
H
H
N
Ph
H
H
H
Ph
H
H
Ph
3a
3b
3c
3
Figure 6. Three conformations about the N–C bond rotation for ligand 3.

490
M.-C. Wang et al. / Tetrahedron: Asymmetry 21 (2010) 486–493
Table 2
The relationship between the free ligand conformational equilibria and the observed enantioselectivity in the addition of diethylzinc to benzaldehyde
Ph
H
R
O
CPh2OH
H
CPh2OH
H
H
R
H
CPh2OH
H
H
R
H
N
Ph
R1
N
R2
N
R3
N
R
H
R2
R3
R3
R3
R2
R1
R1
R2
R1
c
b
a
Ratiob (%)
ee
a
Entry
Ligands
R
Comf
E (Hartree)
Erel (kJ/mol)
1
2
3
4
5
6
7
8
9
3 (R1 = Ph, R2 = R3 = H)
4 (R1 = Ph, R2 = R3 = H)
5 (R1 = R2 = Ph, R3 = H)
6 (R1 = R2 = R3 = Ph)
H
H
H
Me
Me
Me
H
H
H
3a
3b
3c
4a
4b
4c
5a
5b
5c
6
ꢀ980.97938
ꢀ980.97832
ꢀ980.97773
ꢀ1020.29807
ꢀ1020.29716
ꢀ1020.29319
ꢀ1212.04074
ꢀ1212.03674
ꢀ1212.03907
0.0
2.8
4.3
0.0
2.4
12.8
0.0
10.5
4.4
66.8
21.4
11.8
72.2
27.4
0.30
84.5
1.20
14.3
100
49.015a
75.015a
96.09b
99.015b
10
H
a
Relative to the lowest energy conformation (shown as zero kJ/mol).
Conformational population is obtained from a Boltzmann distribution of relative energies at 298 K.
b
Ph
Ph
Ph
H
H
H
H
H
Ph
Ph
H
O
H
H
N
Ph
Ph
Ph
Me
Me
Me
HO
HO
H
HO
Ph
H
H
Me
H
N
N
N
Ph
H
H
H
Ph
H
Ph
H
4c
4a
4
4b
Figure 7. Three conformations about the N–C bond rotation for ligand 4.
enantioselectivity (75% ee) for 4 than that (49% ee) for 3. This result
shows that the introduction of methyl group did not affect the flex-
ibility of the backbone, yet it resulted in the equilibrium redistribu-
tions of the conformational isomers. Consequently, the increase in
the population of the desired conformation 4b led to an improve-
ment in reaction enantioselectivity.
dryl group had two desired conformations 5a and 5b (i.e., 85.7% pop-
ulations for the sum of two desired conformations 5a and 5b)
because conformations 5a and 5b, respectively, had a phenyl substi-
tuent that was oriented in the same direction as the axial phenyl
group, as in 5a and 5b. Therefore, both 5a and 5b favored the
improvement of the reaction enantioselectivity. As a result, the chi-
ral ligand 5 gave enantioselectivities of up to 96% ee.
Ligand 5 (Fig. 8),9b which contains a gem-diphenyl group on the
nitrogen atom of the three-membered ring adopted a preferred
conformation 5a in order to avoid strong interaction between the
aziridine ring unit and phenyl group in 5b and 5c, respectively.
The crystal structure of 5 also clearly illustrated this arrange-
ment.9b The theoretical calculations showed that the preferred con-
formation 5a constituted of about 84.5% of the equilibrium mixture
(Table 2, entry 7). Unlike ligands 3 and 4 which contain only one
desired conformation 3b, or 4b, respectively, ligand 5 with a benzhy-
Ligand 6 (Fig. 9)15b has only a single conformation because a
Ph3C substituent on the nitrogen atom displays the expected
C3-symmetry in its substituent conformational preference, that is,
due to the presence of a quaternary center that is usually used to
control the conformation,15 there is always one phenyl group
pointing toward the same direction of the axial phenyl group on
the diphenylhydroxymethyl group with respect to the five-mem-
bered ring, as indicated in 6. Therefore, ligand 6 with one single
Ph
Ph
H
Ph
H
H
Ph
O
H
H
N
Ph
Ph
Ph
Ph
HO
Ph
HO
HO
Ph
H
N
H
H
H
N
N
Ph
Ph
H
Ph
Ph
Ph
H
Ph
5a
5b
5c
5
Figure 8. Three conformations about the N–C bond rotation for ligand 5.

M.-C. Wang et al. / Tetrahedron: Asymmetry 21 (2010) 486–493
Ph
491
OH
O
ZnEt2 , 5% mol 9
H
Ph
Ph
O
Toluene, 0-5 oC, rt, 48h
Ph
Up to 98.5% ee
H
H
Ph
H
H
N
Ph
HO
Ph
H
Scheme 2. Addition of diethylzinc to benzaldehyde.
Ph
Ph
N
Ph
Ph
Ph
3. Conclusion
6
6
In conclusion, we have shown that a relationship must exist be-
tween the free ligand conformational populations and the ob-
served ee values in the asymmetric addition of diethylzinc to
benzaldehyde. The increase in populations of the desired free li-
gand conformations led to an enhancement of the reaction enanti-
oselectivity. Based on this necessary correlation between free
ligand conformational populations and the observed ee values, it
is an alternative simple and practical method that the concept of
conformational design is used to develop efficient chiral ligands
in catalytic asymmetric reactions by means of knowledge of con-
formational analysis. Studies are currently underway using the
concept of conformational design to synthesize new and efficient
chiral ligands for asymmetric catalysis.
Figure 9. Conformation about the N–C bond rotation for ligand 6.
conformation (i.e., 100% desired conformation) gave the best asym-
metric induction of up to 99% ee.
Based on the aforementioned relationship between the free li-
gand conformational populations and the observed ee values,
new compound 9 (Fig. 10) with only one single conformation
should also be the same outstanding chiral ligand as 6 because
the introduction of the methyl group, compared with the ligand
6, cannot alter the equilibrium redistributions of the conforma-
tional isomers.
To further verify this prediction, the new chiral ligand 9 was
synthesized in enantiopure form from starting material allo-L-thre-
onine using the protocol outlined in Scheme 1.9a,16
4. Experimental
4.1. General
With chiral ligand 9 in hand, the asymmetric addition of dieth-
ylzinc to benzaldehyde was examined in toluene in 0–5 °C in the
presence of 5% ligand 9. As expected, ligand 9 with a single confor-
mation gave the desired product with outstanding enantioselectiv-
ities of up to 98.5% ee (Scheme 2). For the sake of comparing the
asymmetric induction efficiency of 6 and 9, the chiral ligand 6
was also tested under the same conditions as 9, and the ligand 6
afforded 98.7% ee. From these results, the ligand 9 did not result
in a change in the enantioselectivity (relative to 6) because there
was no change in the equilibrium distributions of the conforma-
tional isomers even when a methyl group was introduced on the
aziridine ring. However, looking at 2 versus 1 and 4 versus 3, the
introduction of a methyl group led to a substantial increase in
the reaction enantioselectivity because the introduction of a
methyl group resulted in the equilibrium redistributions of the
conformational isomers.
Unless otherwise noted, all reactions were carried out under ar-
gon or nitrogen using standard Schlenk and vacuum line tech-
niques. Toluene was freshly distilled over calcium hydride prior
to use. Other reagents were obtained from commercial sources
and used as received without further purification. Melting points
were determined using YRT-3 melting point apparatus and are
uncorrected. Optical rotations were measured with Perkin–Elmer,
model 341 Polarimeter at 20 °C in CHCl3. The enantiomeric purity
was determined by HPLC using a chiral column with hexane/pro-
pan-2-ol (ratio as indicated) as the eluent. The chromatographic
system consisted of a JASCO model PU-1580 intelligent HPLC pump
and a JASCO model UV-1575 intelligent UV–vis detector (254 nm).
The injection loop had a 20 lL capacity. The column used was a
Chiralcel OD (250 ꢁ 4.6 mm) from Daicel Chemical Ind., Ltd (Ja-
pan). The column was operated at ambient temperature. NMR
spectra (1H and 13C) were performed on a Bruker DPX 400
(400 MHz) spectrometer using solutions in CDCl3 (referenced
internally to Me4Si); J values are given in Hz. TLC was performed
on dry silica gel plates developed with hexane/ethyl acetate. Mass
spectra were obtained using a Bruker esquire-3000 instrument
with an electrospray ionization source (ESIMS). All the ESIMS spec-
tra were performed using MeOH as the solvent. Methanol was
dried with Mg(OCH3)2.
Ph
H
H
Ph
Ph
O
H
H
N
Ph
Me
HO
Ph
H
Me
Ph
Ph
N
Ph
Ph
Ph
9
9
4.2. Synthesis of chiral compound 8
Figure 10. Conformation about the N–C bond rotation for ligand 9.
To a stirred suspension of allo-L-threonine methyl ester hydro-
chloride (2.5 g, 15 mmol) in dichloromethane (20 mL), triethyl-
amine (4.2 mL, 30 mmol) was added dropwise at 0 °C, and then
trityl chloride (4.18 g, 15 mmol) dissolved in chloroform was
added dropwise. After stirring for 72 h at 0 °C, the mixture was
washed with 10% aqueous citric acid solution (3 ꢁ 10 mL) and
water (3 ꢁ 10 mL). After the combined organic extracts were dried
over Na2SO4, the solvent was removed in vacuo. The residue was
purified by recrystallization (ethyl acetate/petroleum) to afford
the product 7 as a yellowish solid (4.7 g, 84%), mp145–147 °C.
Ph
CO2CH3
C6H5MgX
OH
Ph
CO2CH3
MeSO2Cl
NEt3, THF
H3C
OH
H3C
H3C
N
N
NHCPh3
CPh3
CPh3
9: 91%
7
8: 76%
½
a 2D0
ꢂ
¼ þ17:1 (c 1.5, CHCl3). 1H NMR (400 MHz, CDCl3) d 1.11 (d,
Scheme 1. Synthesis of chiral ligand 9.
J = 6.8 Hz, 3H, CH3), 2.40 (br, 1H), 3.01 (d, J = 9.6 Hz, 1H, CH–CO2),

492
M.-C. Wang et al. / Tetrahedron: Asymmetry 21 (2010) 486–493
3.23 (s, 3H, OCH3), 3.38–3.41 (m, 1H, CH3CH), 3.99 (br, 1H), 7.17–
7.28 (m, 9H, PhH), 7.49–7.51 (m, 6H, PhH). 13C NMR: (100 MHz,
CDCl3) d 19.5, 51.5, 61.1, 69.9, 70.9, 126.5, 127.8, 128.8, 145.6,
173.1. IR (KBr): 3512, 3361, 3068, 3050, 2970, 2936, 1713, 1595,
1490, 1444, 1429, 1266, 1223, 1090, 1050, 1090, 778, 751, 705.
MS (ESI): m/z (M + Na) + calcd for C24H25NO3: 389.5; found: 398.6.
sure. Purification of the residue by the preparative silica gel TLC
plate (hexane/EtOAc = 4:1) afforded the (S)-1-phenyl-1-propanol.
The ee was determined by HPLC analyses using a chiral column
(a Chiralcel OD). Hexane/i-PrOH = 100:2, 1 mL/min, tR = 13.2 min,
tS = 16.2 min.
To a stirred solution of trityl allo-L-threonine methyl ester 7
Acknowledgments
(1.3 g, 3.47 mmol) in tetrahydrofuran (9.8 mL), triethylamine
(6.9 mL, 50 mmol) was added dropwise at 0 °C. Methanesulfonyl
chloride (0.4 mL, 5.2 mmol) was added dropwise. After stirring
for 30 min at room temperature, the solution was refluxed for an-
other 48 h. The solvent was removed in vacuo to leave a residue
which was taken up in ethyl acetate (8 mL) and washed with
10% aqueous citric acid solution (3 ꢁ 5 mL) followed by saturated
aqueous sodium bicarbonate solution (2 ꢁ 5 mL). After the com-
bined organic extracts were dried over Na2SO4, the solvent was re-
moved in vacuo. The residue was purified on a preparative silica
gel TLC plate (petroleum/EtOAc = 20:1) and afforded compound 8
(0.94 g) in 76% yield, mp 140–142.5 °C, (lit.17 mp 141–143 °C).
We are grateful to the National Natural Sciences Foundation of
China (NNSFC: 20672102) and Henan Outstanding Youth Program
(084100410001).
References
1. (a)Catalytic Asymmetric Synthesis; Ojima, I., Ed.; Wiley: New York, 2000;
(b)Comprehensive Asymmetric Catalysis; Jacobsen, E. N., Pfaltz, A., Yamamoto, H.,
Eds.; Springer: Berlin, 1999; (c)New Frontiers in Asymmetric Catalysis; Mikami,
K., Lautens, M., Eds.; Wiley: New Jersey, 2007.
2. For reviews, see: (a) Shimizu, K. D.; Snapper, M. L.; Hoveyda, A. H. Chem. Eur. J.
1998, 4, 1885; (b) Gennari, C.; Piarulli, U. Chem. Rev. 2003, 103, 3071; (c) Ding,
K.; Du, H.; Yuan, Y.; Long, J. Chem. Eur. J. 2004, 10, 2872; (d) Jäkel, C.; Paciello, R.
Chem. Rev. 2006, 106, 2912.
3. For reviews, see: Balcells, D.; Maseras, F. New J. Chem. 2007, 31, 333.
4. (a) Harriman, D. J.; Lambropoulos, A.; Deslongchamps, G. Tetrahedron Lett.
2007, 48, 689; (b) Deeth, R. J.; Fey, N. Organometallics 2004, 23, 1042; (c)
Moitessier, N.; Chrètien, F.; Chapleur, Y.; Maigret, B. Eur. J. Org. Chem. 2000,
995; (d) Moitessier, N.; Henry, C.; Len, C.; Chapleur, Y. J. Org. Chem. 2002, 67,
7275; (e) Fristrup, P.; Jensen, G. H.; Andersen, M. L. N.; Tanner, D.; Norrby, P. O.
J. Organomet. Chem. 2006, 691, 2182; (f) Bernardi, A.; Gennari, C.; Goodman, J.
M.; Paterson, I. Tetrahedron: Asymmetry 1995, 6, 2613; (g) Corbeil, C. R.;
Thielges, S.; Schwartzentruber, J. A.; Moitessier, N. Angew. Chem., Int. Ed. 2008,
47, 2635.
½
a 2D0
a 2D0
ꢂ
¼ þ11:3 (c 1.38, CHCl3), {lit.17
½
a 2D0
ꢂ
¼ þ10 (c 1.0, CHCl3)}.
½
ꢂ
¼ þ11:3 (c 1.38, CHCl3). 1H NMR (400 MHz, CDCl3): d 0.60
(d, J = 6.4 Hz, 3H, CH3), 2.35 (d, J = 2.4 Hz, 1H, CH–CO2), 2.96–3.01
(m, 1H, CH3CH), 3.70 (s, 3H, OCH3), 7.19–7.30 (m, 9H, PhH),
7.49–7.51 (m, 6H, PhH). IR (KBr): 3060, 3021, 2952, 1748, 1596,
1489, 1445, 1274, 1199, 1178, 1065, 1032, 755, 706. MS (ESI): m/
+
z (M+Na) calcd for C24H23NO2: 380.4; found: 379.9.
4.3. Synthesis of chiral compound 9
5. (a) Ianni, J. C.; Annamalai, V.; Phuan, P.; Panda, M.; Kozlowski, M. C. Angew.
Chem., Int. Ed. 2006, 45, 5502; (b) Huang, J.; Ianni, J. C.; Antoline, J. E.; Hsung, R.
P.; Kozlowski, M. C. Org. Lett. 2006, 8, 1565; (c) Kozlowski, M. C.; Waters, S. P.;
Skudlarek, J. W.; Evans, C. A. Org. Lett. 2002, 4, 4391; (d) Vidal-Ferran, A.;
Moyano, A.; Pericàs, M. A.; Riera, A. Tetrahedron Lett. 1997, 38,
8773.
6. Reviews on enantioselective organozinc additions to aldehydes: (a) Noyori, R.;
Kitamura, M. Angew. Chem., Int. Ed. Engl. 1991, 30, 49; (b) Soai, K.; Niwa, S.
Chem. Rev. 1992, 92, 833; (c) Pu, L.; Yu, H.-B. Chem. Rev. 2001, 101, 757; (d)
Seebach, D.; Beck, A. K.; Heckel, A. Angew. Chem., Int. Ed. 2001, 40, 92; (e) Walsh,
P. Acc. Chem. Res. 2003, 36, 739; (f) Zhu, H.-J.; Jiang, J.-X.; Ren, J.; Yan, Y.-M.;
Pittman, C. U. Cur. Org. Synth. 2005, 2, 547; (g) Hatano, M.; Miyamoto, T.;
Ishihara, K. Cur. Org. Chem. 2007, 11, 127.
A Grignard reagent was prepared in the usual way from 146 mg
(6 mmol) of magnesium and bromobenzene 6 mmol in THF (5 mL).
The solution was cooled to ꢀ20 °C before the addition of a solution
of 8 (515 mg, 1.5 mmol) in THF (2 mL). The mixture was allowed to
reach the room temperature. After stirring for 24 h, the reaction
was quenched with saturated aqueous NH4Cl (10 mL) at 0 °C. The
phases were separated and the aqueous phase was extracted with
Et2O (3 ꢁ 5 mL). The combined organic phases were washed with
brine (15 mL), dried over Na2SO4, and after filtration the solvent
was removed under reduced pressure. The resulting residue was
purified by the preparative TLC with petroleum/EtOAc (20:1) as
developing solvent to give 9 (665 mg, 91%), mp 164.5–165.9 °C.
7. Zhu, H. J.; Jiang, J. X.; Saebo, S.; Pittman, C. U., Jr. J. Org. Chem. 2005, 70,
261.
8. (a) Wang, M.-C.; Wang, D.-K.; Zhu, Y.; Liu, L.-T.; Guo, Y.-F. Tetrahedron:
Asymmetry 2004, 15, 1289; (b) Wang, M.-C.; Liu, L.-T.; Zhang, J.-S.; Shi, Y.-Y.;
Wang, D.-K. Tetrahedron: Asymmetry 2004, 15, 3853; (c) Wang, M.-C.; Hou, X.-
H.; Xu, C.-L.; Liu, L.-T.; Li, G.-L.; Wang, D.-K. Synthesis 2005, 3620; (d) Wang, M.-
C.; Zhao, W. X.; Wang, X.-D.; Song, M. P. Synlett 2006, 3443; (e) Wang, M.-C.;
Hou, X.-H.; Chi, C.-X.; Tang, M.-S. Tetrahedron: Asymmetry 2006, 17, 2126; (f)
Wang, M.-C.; Zhang, Q.-J.; Zhao, W. X.; Wang, X.-D.; Ding, X. J. Org. Chem. 2008,
73, 168; (g) Wang, M.-C.; Wang, X.-D.; Ding, X.; Liu, Z.-K. Tetrahedron 2008, 64,
2559; (h) Wang, M.-C.; Zhang, Q.-J.; Li, G.-W.; Liu, Z.-K. Tetrahedron: Asymmetry
2009, 20, 288; (i) Niu, J.-L.; Wang, M.-C.; Lu, L.-J.; Ding, G.-L.; Lu, H.-J.; Chen, Q.-
T.; Song, M.-P. Tetrahedron: Asymmetry 2009, 20, 2616.
½
a 2D0
ꢂ
¼ þ44:2 (c 1.01, CHCl3). 1H NMR (400 MHz, CDCl3): d 0.77
(d, J = 6.4 Hz, 3H, CH3), 2.46 (d, J = 3.2 Hz, 1H, CH3CHCH), 2.65–
2.71 (m, 1H, CH3CH), 3.74 (s, 1H, OH), 6.98–7.31 (m, 25H, PhH).
13C NMR (100 MHz, CDCl3): d 13.54, 37.41, 50.02, 72.53, 74.26,
125.94, 126.15, 126.58, 126.96, 127.40, 127.47, 127.93, 128.05,
129.86, 145.60, 147.18. IR (KBr): 3368, 3084, 3055, 3021, 2956,
2928, 1595, 1489, 1446, 1369, 1334, 1182, 1151, 1069, 1033,
745, 7012. MS (ESI): m/z (M+H)+ calcd for C35H31NO: 482.6; found:
482.5. Anal. Calcd for C35H31NO requires: C, 87.28; H, 6.49; N, 2.91.
Found: C, 87.32; H, 6.47; N, 2.97.
9. (a) Willems, J. G. H.; Hersmis, M. C.; de Gelder, R.; Smits, J. M. M.; Hamink, J. B.;
Dommerholt, F. J.; Thijs, L.; Zwanenburg, B. J. Chem. Soc., Perkin Trans. 1 1997,
963; (b) Bulman Page, P. C.; Allin, S. M.; Maddocks, S. J.; Elsegood, M. R. J. J.
Chem. Soc., Perkin Trans. 1 2002, 36, 2827.
10. Frisch, J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.;
Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.;
Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.;
Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo,
C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G.
A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A.
D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.;
Gonzalez, C.; Pople, J. A. GAUSSIAN 03, Revision A.1; Gaussian, Inc.: Pittsburgh,
PA, 2003.
4.4. General procedure for the asymmetric addition of
diethylzinc arylaldehydes
A solution of diethylzinc (1 M in n-hexane, 1.1 mL) was added
to a solution of a chiral catalyst 9 (0.025 mmol, 5 mol %) in dry tol-
uene under a nitrogen atmosphere. The mixture was cooled to 0 °C
and stirred for 30 min. Freshly distilled benzaldehyde (0.05 mL,
0.5 mmol) was added to the mixture. The resulting mixture was
stirred for 10 h in 0–5 °C and was allowed to warm to room tem-
perature, and kept stirring for another 38 h at the same tempera-
ture. The reaction was quenched by the addition of saturated
aqueous NH4Cl (4 mL). The mixture was extracted with Et2O
(3 ꢁ 8 mL). The combined organic layers were washed with brine,
dried over anhydrous Na2SO4, and evaporated under reduced pres-
11. (a) Chang, Y.-P.; Su, T.-M.; Li, T.-W.; Chao, I. J. Phys. Chem. A 1997, 101, 6107; (b)
Reany, O.; Goldberg, I.; Abramson, S.; Golender, L.; Ganguly, B.; Fuchs, B. J. Org.
Chem. 1998, 63, 8850; (c) Star, A.; Goldberg, I.; Lemcoff, N. G.; Fuchs, B. Eur. J.

M.-C. Wang et al. / Tetrahedron: Asymmetry 21 (2010) 486–493
493
Org. Chem. 1999, 2033; (d) Ganguly, B.; Freed, D. A.; Kozlowski, M. C. J. Org.
Chem. 2001, 66, 1103; (e) Xu, Z.; Kozlowski, M. C. J. Org. Chem. 2002, 67, 3072.
12. (a) Yamakawa, M.; Noyori, R. J. Am. Chem. Soc. 1995, 117, 6327; (b) Yamakawa,
M.; Noyori, R. Organometallics 1999, 18, 128.
13. Rasmussen, T.; Norrby, P.-O. J. Am. Chem. Soc. 2003, 125, 5130.
14. (a) Hoffmann, R. W. Angew. Chem., Int. Ed. 2000, 39, 2054; (b) Hoffmann, R. W.
Angew. Chem., Int. Ed. 1992, 31, 1124.
15. (a) Tanner, D.; Kornø, H. T.; Guijarro, D.; Andersson, P. G. Tetrahedron 1998, 54,
14213; (b) Lawrence, C. F.; Nayak, S. K.; Thijs, L.; Zwanenburg, B. Synlett 1999,
1571.
16. (a) Anderson, N. H.; Bowman, J.; Erwin, A.; Harwood, E.; Kline, T.; Mdluli, K.;
Ng, S.; Pfister, K. B.; Shawar, R.; Wagman, A.; Yabannavar, A. WO2004062601
A2, July 29, 2004.; (b) McKeever, B.; McKeever, G. Tetrahedron 2003, 59, 2713.
17. Dugave, C.; Mfènez, A. Tetrahedron: Asymmetry 1997, 8, 1453.