Highly enantio- and diastereoselective mannich reactions of chiral Ni(II) glycinates with amino sulfones. Efficient asymmetric synthesis of aromatic αβ-diamino acids

Article abstract of DOI:10.1021/jo8019169

(Chemical Equation Presented) This paper describes a practical, enantio- and diastereoselective Mannich reaction between a chiral Ni(II) complex of glycine 1 and α-amino sulfones 2, involving the creation of a carbon-carbon bond and two stereogenic centers in a single operation; it represents an attractive route to the synthesis α,β-diamino acids.

Full text of DOI:10.1021/jo8019169

Highly Enantio- and Diastereoselective Mannich Reactions of Chiral
Ni(II) Glycinates with Amino Sulfones. Efficient Asymmetric
Synthesis of Aromatic r,ꢀ-Diamino Acids
Jiang Wang,† Ting Shi,† Guanghui Deng, Hualiang Jiang, and Hong Liu*
Drug DiscoVery and Design Centre, State Key Laboratory of Drug Research, Shanghai Institute of Materia
Medica, Shanghai Institutes for Biological Sciences, Graduate School, Chinese Academy of Sciences,
555 Zuchongzhi Road, Shanghai 201203, China
hliu@mail.shcnc.ac.cn
ReceiVed September 4, 2008
This paper describes a practical, enantio- and diastereoselective Mannich reaction between a chiral Ni(II)
complex of glycine 1 and R-amino sulfones 2, involving the creation of a carbon-carbon bond and two
stereogenic centers in a single operation; it represents an attractive route to the synthesis R,ꢀ-diamino
acids.
Introduction
nitro moiety to an amino group. Willis et al. meanwhile
described a direct catalytic Mannich reaction that leads to anti-
R,ꢀ-diamino acids.6 However, the removal of the para-tolu-
enesulfonate (tosyl) protecting group requires harsh reaction
conditions. A chiral phase-transfer catalyst was recently dis-
closed by Shibasaki et al. to achieve highly diastereoselective
access to syn-R,ꢀ-diamino acids,7a and Ooi et al. also reported
the asymmetric synthesis of R,ꢀ-diamino acid derivatives under
asymmetric phase transfer conditions.7b Hayashi et al. have
reported asymmetric gold(I)-catalyzed reactions of methyl
isocyanoacetate with imines as potentially useful access to
erythro-R,ꢀ-diamino acids.8 However, these approaches rely on
the use of relatively exotic and therefore expensive catalysts
that might limit their broad synthetic applications. Finally,
Soloshonok et al. has reported a single example of highly
diastereoselective aza-aldol reaction of chiral Ni(II) complex
R,ꢀ-Diamino acids are versatile building blocks in organic
synthesis and medicinal and peptide/peptidomimetic chemistry.1,2
Furthermore, they are also useful in wide applications, as chiral
auxiliaries and ligands for asymmetric synthesis.3 Accordingly,
the development of efficient methods in their preparation has
been a mainstay in organic synthesis.4 Jørgensen et al. report
an indirect method using a chiral copper complex-promoted aza-
Henry reaction of nitro compounds with R-imino esters.5 This
approach requires an additional reduction step, to convert the
† These authors contributed equally to this work.
(1) A review on R,ꢀ-diamino acids: Viso, A.; Ferna´ndez de la Pradillla, R.;
Garc´ıa, A.; Flores, A. Chem. ReV. 2005, 105, 3167.
(2) (a) Stragies, R.; Osterkamp, F.; Zischinsky, G.; Vossmeyer, D. H.; Kalkho,
Reimer, U.; Zahn, G. J. Med. Chem. 2007, 50, 3786. (b) Subasinghe, N.; Schulte,
M.; Roon, R. J.; Koerner, J. F.; Johnson, R. L. J. Med. Chem. 1992, 35, 4602.
(c) Wang, M.; Gould, S. J. J. Org. Chem. 1993, 58, 5176. (d) Lee, E. S.; Jurayj,
J.; Cushman, M. Tetrahedron 1994, 50, 9873. (e) Rossi, F. M.; Powers, E. T.;
Yoon, R.; Rosenberg, L.; Meinwald, J. Tetrahedron 1996, 52, 10279.
(3) Lucet, D.; Le Gall, T.; Mioskowski, C. Angew. Chem., Int. Ed. 1998,
37, 2580.
(4) (a) Davis, F. A.; Zhang, Y.; Qiu, H. Org. Lett. 2007, 9, 833. (b) Davis,
F. A.; Deng, J. Org. Lett. 2004, 6, 2789. (c) Chowdari, N. S.; Ahmad, M.;
Albertshofer, K.; Tanaka, F.; Barbas, C. F. Org. Lett. 2006, 8, 2839. (d) Viso,
A.; Ferna´ndez de la Pradillla, R.; Lo´pez-Rodr´ıguez, M. L.; Garc´ıa, A.; Flores,
A.; Alonso, M. J. Org. Chem. 2004, 69, 1542. (e) Singh, A.; Johnston, J. N.
J. Am. Chem. Soc. 2008, 130, 5866. (f) Chen, Z.; Morimoto, H.; Matsunaga, S.;
Shibasaki, M. J. Am. Chem. Soc. 2008, 130, 2170.
(5) Nishiwaki, N.; Knudsen, K. R.; Gothelf, K. V.; Jørgensen, K. A. Angew.
Chem., Int. Ed. 2001, 40, 2992.
(6) Cutting, G. A.; Stainforth, N. E.; John, M. P.; Kociok-Ko¨hn, G.; Willis,
M. C. J. Am. Chem. Soc. 2007, 129, 10632.
(7) (a) Okada, A.; Shibuguchi, T.; Ohshima, T.; Masu, H.; Yamaguchi, K.;
Shibasaki, M. Angew. Chem., Int. Ed. 2005, 44, 4564. (b) Ooi, T.; Kameda, M.;
Fujii, J.; Maruoka, K. Org. Lett. 2004, 6, 2397.
(8) (a) Soloshonok, V. A.; Hayashi, T. Tetrahedron: Asymmetry 1994, 5,
1091. (b) Hayashi, T.; Kishi, E.; Soloshonok, V. A.; Uozumi, Y. Tetrahedron
Lett. 1996, 37, 4969.
10.1021/jo8019169 CCC: $40.75
Published on Web 10/10/2008
2008 American Chemical Society
J. Org. Chem. 2008, 73, 8563–8570 8563

Wang et al.
SCHEME 1. Stereoselective Synthesis of Enantioenriched r,ꢀ-Diamino Acids via Chiral Ni(II) Complexes with Sulfones
of glycine (S)-1 with extremely electrophonic trifluoroacetaldi-
mine as a plausible method for preparation of 3-perfluoroalkyl-
2,3-diamino acids.9 Thus, various methodological and structural
shortcomings of the literature approaches render these methods
of limited synthetic application, in particular for preparation of
the target diamino acids on relatively large scale. In this paper,
we disclose a strategy for the practical synthesis of highly
enantio- and diastereoselective R,ꢀ-diamino acids, with high
efficiency. Significantly, the process developed in this work
features rare synthetic methods for operational convenience10
and is based on readily available and inexpensive starting
materials.
aldol,11 Michael addition,12 and C-alkylation reactions.13 The
notable merits of the Ni(II) complex’s methods are (1) the
provision of high enantio- and/or diastereocontrol, (2) the use
of readily available and cost-effective substances, (3) mild
reaction conditions and simple reaction procedures, (4) high
reaction yields with good reproducibility, and (5) easy recovery
of chiral ligands. These unique properties render it an attractive
strategy for practical synthesis in industrial settings.14 Accord-
ingly, we proposed the use of the chiral Ni(II) complex as a
stereocontroller for the asymmetric synthesis of R,ꢀ-diamino
acids. Remarkably, we demonstrated that the Mannich reaction
between a chiral Ni(II) complex of glycine 1 and R-amino
sulfones 2 served as an efficient approach to R,ꢀ-diamino acids,
with the creation of a new carbon-carbon bond and two
adjacent stereogenic centers in a single operation (Scheme 1).
This is an example of highly stereoselective Mannich reactions
with an in situ generation of carbamate-protected imines by
chiral Ni(II) complex. As demonstrated, robust and readily
obtained Boc/Cbz-R-amino sulfones 215 with wide ranges of
structural diversity can efficiently participate in the process with
high enantio- and diastereoselectivity, thus significantly expand-
ing the scope of the reactions. Furthermore, the use of readily
manipulated Boc/Cbz protecting groups enables the use of these
highly enantioenriched amino acid products in further synthetic
elaborations.
The Ni(II) complex of the chiral Schiff base of glycine has
been widely used to synthesize enantiopure R-amino acids via
(9) Soloshonok, V. A.; Avilov, D. V.; Kukhar, V. P.; Meervelt, L. V.;
Mischenko, N. Tetrahedron Lett. 1997, 38, 4671.
(10) (a) Boyall, D.; Frantz, D. E.; Carreira, E. M. Org. Lett. 2002, 4, 2605.
(b) Soloshonok, V. A.; Ohkura, H.; Sorochinsky, A.; Voloshin, N.; Markovsky,
A.; Belik, M.; Yamazaki, T. Tetrahedron Lett. 2002, 43, 5445. (c) Sorochinsky,
A.; Voloshin, N.; Markovsky, A.; Belik, M.; Yasuda, N.; Uekusa, H.; Ono, T.;
Berbasov, D. O.; Soloshonok, V. A. J. Org. Chem. 2003, 68, 7448. (d)
Soloshonok, V. A.; Berbasov, D. O. J. Fluorine Chem. 2004, 125, 1757. (e)
Moore, J. L.; Taylor, S. M.; Soloshonok, V. A. ARKIVOC 2005, 6, 287. (f)
Yasumoto, M.; Ueki, H.; Soloshonok, V. A. J. Fluorine Chem. 2007, 128, 736.
(11) (a) Belokon, Y. N.; Bulychev, A. G.; Vitt, S. V.; Struchkov, Y. T.;
Batsanov, A. S.; Timofeeva, T. V.; Tsyryapkin, V. A.; Ryzhov, M. G.; Lysova,
L. A.; Bakhmutov, V. I.; Belikov, V. M. J. Am. Chem. Soc. 1985, 107, 4252.
(b) Soloshonok, V. A.; Kukhar, V. P.; Galushko, S. V.; Svistunova, N. Y.; Avilov,
D. V.; Kuzmina, N. A.; Raevski, N. I.; Struchkov, Y. T.; Pisarevsky, A. P.;
Belokon, Y. N. J. Chem. Soc., Perkin Trans. 1 1993, 3143. (c) Soloshonok,
V. A.; Avilov, D. V.; Kukhar′, V. P. Tetrahedron: Asymmetry 1996, 7, 1547.
(d) Soloshonok, V. A.; Avilov, D. V.; Kukhar, V. P. Tetrahedron 1996, 52,
12433.
Results and Discussion
Mannich Reactions between (S)-Ni(II) Complex 1 and
N-(tert-Butoxycarbonyl)-(-)-(phenylsulfonyl)benzylamine 2a
Using Various Bases and Solvents. On the basis of the
successful synthesis of R-amino-ꢀ-hydroxy acids,11 we selected
chiral (S)-Ni(II) complex of glycine 1 with (S)-o-[N-(N-
(12) (a) Soloshonok, V. A.; Cai, C.; Hruby, V. J. Angew. Chem., Int. Ed.
2000, 39, 2172. (b) Soloshonok, V. A.; Cai, C.; Hruby, V. J. Tetrahedron Lett.
2000, 41, 9645. (c) Cai, C.; Soloshonok, V. A.; Hruby, V. J. J. Org. Chem.
2001, 66, 1339. (d) Soloshonok, V. A.; Cai, C.; Yamada, T.; Ueki, H.; Ohfune,
Y.; Hruby, V. J. J. Am. Chem. Soc. 2005, 127, 15296. (e) Soloshonok, V. A.;
Cai, C.; Hruby, V. J. Tetrahedron Lett. 2000, 41, 135.
(13) (a) Gu, X.; Tang, X.; Cowell, S.; Ying, J.; Hruby, V. J. Tetrahedron
Lett. 2002, 43, 6669. (b) Taylor, S. M.; Yamada, T.; Ueki, H.; Soloshonok, V. A.
Tetrahedron Lett. 2004, 45, 9159. (c) Belokon, Y. N.; Sagyan, A. S.; Djamgaryan,
S. M.; Bakhmutov, V. I.; Belikov, V. M. Tetrahedron 1988, 44, 5507. (d)
Soloshonok, V. A.; Tang, X.; Hruby, V. J.; Meervelt, L. V. Org. Lett. 2001, 3,
341. (e) Soloshonok, V. A.; Yamada, T.; Ueki, H.; Moore, A. M.; Cook, T. K.;
Arbogast, K. L.; Soloshonok, A. V.; Martin, C. H.; Ohfune, Y. Tetrahedron
2006, 62, 6412.
(14) (a) Cai, C.; Yamada, T.; Tiwari, R.; Hruby, V. J.; Soloshonok, V. A.
Tetrahedron Lett. 2004, 45, 6855. (b) Belokon, Y. N.; Kochetkov, K. A.;
Ikonnikov, N. S.; Strelkova, T. V.; Harutyunyan, S. R.; Saghiyan, A. S.
Tetrahedron: Asymmetry 2001, 12, 481.
(15) (a) Fini, F.; Sgarzari, V.; Pettersen, D.; Herrera, R. P.; Bernardi, L.;
Ricci, A. Angew. Chem., Int. Ed. 2005, 44, 7975. (b) Song, J.; Shih, H.; Deng,
L. Org. Lett. 2007, 9, 603.
8564 J. Org. Chem. Vol. 73, No. 21, 2008

Enantio- and DiastereoselectiVe Mannich Reactions
TABLE 1. Optimization of the Reaction Conditionsa
TABLE 2. Asymmetric Mannich Reactions of (S)-Ni(II) Complex
1 with r-Amino Sulfones 2a
entry
base
solvent
temp °C yield % syn:antib de %c
entry
product
R group
Ph
2-Me-C6H4
3-Me-C6H4
4-Me-C6H4
4-MeO-C6H4
4-NO2-C6H4
4-F-C6H4
4-Cl-C6H4
4-Br-C6H4
2-furyl
3-thiophenyl
2-naphthyl
yield % syn:antib de %c
1
2
3
4
5
6
7
8
DBU
NaH
tBuOK CH2Cl2
CH2Cl2
CH2Cl2
23
23
86
91
91
92
93
93
92
85
80
85
80
92
96
81:19
65:35
64:36
71:29
90:10
92:8
>99
91
1
2
3
4
5
6
7
8
(S,2S,3R)-3a
(S,2S,3R)-3b
(S,2S,3R)-3c
(S,2S,3R)-3d
(S,2S,3R)-3e
(S,2S,3R)-3f
(S,2S,3R)-3g
(S,2S,3R)-3h
(S,2S,3R)-3i
(S,2S,3R)-3j
(S,2S,3R)-3k
(S,2S,3R)-3l
93
92
93
91
90
91
94
93
94
91
92
90
90
25
16
17
92:8
91:9
92:8
93:7
92:8
93:7
94:6
93:7
94:6
90:10
91:9
92:8
90:10
90:10
91:9
90:10
>99
91
96
>99
94
79
87
73
91
>99
92
>98
88
77
23
90
NaOH
DBU
DBU
DBU
DBU
DBU
DBU
DBU
DBU
DBU
CH2Cl2
CH3CN
acetone
THF
23
87
23
23
23
23
40
0
-20
-40
-60
>99
>99
98
>99
>99
>99
>99
>99
>99
80:20
83:17
76:24
87:13
87:13
90:10
92:8
DMF
9
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
9
10
11
12
13
10
11
12
13
14
15
16
(S,2S,3R)-3m 2-quinoline
a Reactions were run with 0.20 mmol of (S)-1, 0.21 mmol of 2a in 10
mL of solvent with 0.24 mmol base for 4 h. b Determined by HPLC
analysis. c Determined by SFC analysis (see Supporting Information for
details).
(S,2S,3R)-3n
(S,2S,3R)-3o
(S,2S,3R)-3p
cyclohexyl
i-Pr
Et
>99
94
a Reactions were run with 0.20 mmol of (S)-1, 0.21 mmol of 2 in 10
mL of acetone with 0.24 mmol DBU for 4 h under ambient conditions.
b Determined by HPLC analysis. c Determined by SFC analysis (see
Supporting Information for details).
benzylprolyl)amino]benzophenone as a chiral equivalent nu-
cleophilic partner and chose to study the union of (S)-1 and the
R-amino sulfone 2a derived from benzaldehyde as a model
substrate for optimizing reaction conditions. The results are
summarizedinTable1.Intheinitialstudy,1,8-diazabicyclo[5.4.0]-
undec-7-ene (DBU) was selected as a base9 and the reaction
was conducted in dichloromethane. Gratifyingly, the process
delivered the Mannich adduct (S,2S,3R)-3a in high enantiose-
lectivity 99% de (Table 1, entry 1). We used 1.2 equiv of DBU
as base for this Mannich reaction; 1 equiv of DBU was used to
generate in situ the corresponding imines, and the other 0.2 equiv
of DBU was used for purely catalytic purposes.9 A variety of
alternative bases were investigated accordingly. It was found
that a moderate diastereoselectivity (entries 2-4) was observed
with the use of NaH, t-BuOK, and NaOH. Screening of solvents
(entries 5-8) revealed that, while dichloromethane and aceto-
nitrile afforded good results, the best combination of yield and
enantioselectivity was realized with acetone (Table 1, entry 6).
The representative results collected in entries 1-8 show that
the diastereoselectivity greatly depends on the conditions
applied, while the thermodynamically controlled stemochemical
outcome could be only slightly influenced by the nature of both
base and solvent used. Further optimization studies established
that with the reaction with the base DBU and dichloromethane
at various temperatures, from 40 to -60 °C (entries 9-13), in
all cases good yields and excellent enantioselectivities were
achieved. Lowering the reaction temperature led to significant
improvement in the diastereoselectivity. From the viewpoint of
practical applications, we chose DBU as a base and acetone as
a solvent, both at ambient temperature, to probe the generality
of the Mannich process (Table 1, entry 6).
The system was inert to the steric effect. The three regioisomeric
R-amino sulfones 2 effectively participated in the Mannich
reactions while achieving equally high levels of yield and
selectivity (entries 1-4). In general, functionalized aryl imines
are excellent substrates for the reaction, regardless of the
electronic effect. The aromatic system bearing electron-donating
and -withdrawing groups could be tolerated (entries 5-9);
moreover, the process could be applied to heterocyclic com-
pounds (entries 10-13). The less reactive aliphatic substrates
also engaged in the reaction with good to high de values, in
spite of low yields (entries 14-16). The aliphatic products were
obtained with lower yields than the aryl ones, which may be
consistent with the view that the aryl substituents reduce the
electron density on the carbon atom of the R-amino sulfones,
enhancing its reactivity, while aliphatic substituents weaken the
effect. The syn-diastereomer was obtained as the major product,
based on a single X-ray crystal structural analysis (Figure S1
in the Supporting Information).
Mannich Reactions between (R)-Ni(II) Complex 1 and
r-Amino Sulfones 2. The successful preparation of (S)-amino
acids by the electronic si side attack of the Ni(II) complex of
(S)-1 enolate double bond using simple halides was widely
published by Belokon et al.;16 however, the preparation of (R)-
amino acids through the Ni(II) complex of (R)-1 has been rarely
reported. Table 3 summarizes the asymmetric Mannich reactions
of substrate (R)-1 with different R-amino sulfones 2, under
optimized conditions at hand (Table 1, entry 6). All reactions
were conducted under the same reaction conditions, and all
resulted in a virtually quantitative yield and high diastereo-
selelctivity (entries 1-13). The three regioisomeric R-amino
sulfones 2 generated from tolualdehyde all provided the Mannich
Mannich Reactions between (S)-Ni(II) Complex 1 and
r-Amino Sulfones 2. After the reaction conditions were
optimized, the generality of the reaction was investigated, and
the results, summarized in Table 2, show that the reaction has
broad applicability. Examination of the results reveals that the
process served as a general approach to syn-R,ꢀ-diamino acids.
(16) Tararov, V. I.; Savel’eva, T. F.; Kuznetsov, N. Y.; Ikonnikov, N. S.;
Orlova, S. A.; Belokon, Y. N.; North, M. Tetrahedron: Asymmetry 1997, 8, 79.
J. Org. Chem. Vol. 73, No. 21, 2008 8565

Wang et al.
TABLE 3. Asymmetric Mannich Reactions of (R)-Ni(II) Complex 1 with r-Amino Sulfones 2a
entry
product
R group
Ph
yield %
syn:antib
de %c
1
2
3
4
5
6
7
8
(R,2R,3S)-3a
(R,2R,3S)-3b
(R,2R,3S)-3c
(R,2R,3S)-3d
(R,2R,3S)-3e
(R,2R,3S)-3f
(R,2R,3S)-3g
(R,2R,3S)-3h
(R,2R,3S)-3i
(R,2R,3S)-3j
(R,2R,3S)-3k
(R,2R,3S)-3l
(R,2R,3S)-3m
(R,2R,3S)-3n
(R,2R,3S)-3o
(R,2R,3S)-3p
94
93
93
94
95
86
92
94
95
92
90
87
86
23
18
19
92:8
90:10
91:9
93:7
92:8
93:7
93:7
94:6
94:6
91:9
90:10
91:9
90:10
92:8
90:10
91:9
>99
90
89
>99
79
89
88
84
86
80
81
88
94
>99
94
88
2-Me-C6H4
3-Me-C6H4
4-Me-C6H4
4-MeO-C6H4
4-NO2-C6H4
4-F-C6H4
4-Cl-C6H4
4-Br-C6H4
2-furyl
3-thiophenyl
2-naphthyl
2-quinoline
cyclohexyl
i-Pr
9
10
11
12
13
14
15
16
Et
a Reactions were run with 0.20 mmol of (R)-1, 0.21 mmol of 2 in 10 mL of acetone with 0.24 mmol DBU for 4 h under ambient conditions.
b Determined by HPLC analysis. c Determined by SFC analysis (see Supporting Information for details).
SCHEME 2. Decomposition of Ni(II) Complex to Release Amino Acids and Recovery of the Ligand (S)-BPB
adducts with equally high levels of yield and selectivity to the
parent system (entries 1-4). In general, functionalized aryl
imines are excellent substrates for the reaction, with examples
of electron-donating and -withdrawing groups and a number of
halo-substituted examples all performing well (entries 5-9). The
use of 2-furyl-, 3-thiophenyl-, 2-naphthyl-, and 2-quinoline
imines demonstrates that heteroaromatic imines can be suc-
cessfully employed (entries 10-13). The cyclohexyl-derived
imine delivered the Mannich adduct at 23% yield with >99%
de, although the smaller i-propyl- and ethyl-derived imines
showed reduced yields (entries 14-16). In all cases, the syn-
diastereomer was obtained as the major product. The relative
stereochemistry (syn/anti) of the Mannich products was deter-
mined by X-ray analysis of (R,2R,3S)-3a, and the absolute
configuration is shown in Figure S1 in the Supporting Informa-
tion.
Decomposition of Products (S,2S,3R)-3a and (S,2S,3R)-
5a, Isolation of 2,3-Diamino-3-phenylpropanoic Acid (2S,3R)-
4a·HCl and Nꢀ-Cbz-2,3-Diamino-3-phenylpropanoic Acid
(2S,3R)-6a·HCl, and Recovery of the Ligand (S)-BPB. It was
shown that the chiral ligand BPB (Scheme 2) was easily
recovered in quantitative yield and could be reused by using a
simple procedure. The decomposition of compound (S,2S,3R)-
3a under standard conditions by heating a suspension of
(S,2S,3R)-3a in methanol/6 N HCl afforded the target amino
acid 2,3-diamino-3-phenylpropanoic acid (2S,3R)-4a·HCl in
96% yield, and meanwhile (S)-BPB was recovered quantitatively
(Scheme 2). Moreover, we demonstrated that the method can
be scaled in gram-scale synthesis (5 g).
Finally and importantly, it was also shown that the process
could be applied to Cbz-R-amino sulfone as a donor for this
asymmetric Mannich reaction (Scheme 3). The resulting
product (S,2S,3R)-5a was readily converted to Nꢀ-Cbz-2,3-
diamino-3-phenylpropanoic acid (2S,3R)-6a · HCl. The or-
thogonally protected amino acid protecting group in (2S,3R)-
6a · HCl enabled ready manipulation in further synthetic
elaborations.
Quantum Chemical Calculations. Theoretical calculations
have been carried out to understand the reaction mechanism
and the high enantio- and diastereoselectivity. We chose
reactions with (S)-Ni(II) complex and R-amino sulfones as
our investigated subjects. According to our calculations, the
reaction involves four fundamental steps (Scheme 4): (1) the
enolization of (S)-Ni(II) complex 1, promoted by DBU with
a hydrogen bond to form intermediate M; (2) Ni(II)-enolate
reacts with imine obtained from R-amino sulfone 2a, to give
8566 J. Org. Chem. Vol. 73, No. 21, 2008

Enantio- and DiastereoselectiVe Mannich Reactions
SCHEME 3. Synthesis of Nꢀ-Cbz-2,3-Diamino-3-Phenylpropanoic Acid·HCl
SCHEME 4. Proposed Cycle for (S)-Ni(II) Complex Mannich Reaction
transition states of syn-TS and anti-TS; (3) the release of
(2S,3R)-4a · HCl adduct; and (4) the regeneration of (S)-Ni(II)
complex 1.
than (S,2S,3S)-anti-TS. In conclusion, the steric repulsion is
expected to be the essential factor contributing to the
instability of anti-TSs. The computed energy parameters of
these transition states are given in Table 4.
A final step was the release of syn-4a and anti-4a and
regeneration of (S)-Ni(II) complex 1.
The intriguing effect of the DBU motivated us to study
its function; as addressed in the Experimental Section,
reactions employing DBU obtained the Mannich adduct
(2S,3R)-4a · HCl with very high enantioselectivity 99% de
(Table1, entries 1, 5, and 6). It is thought that the nitrogen
of DBU connected with the oxygen of complex 1 by hydrogen
bonding to promote the enolization of complex 1. We then
put them together to optimize the minima; fortunately, our
computational results indicated that the hydrogen bond
(O · · · H · · · N) really exists and that it makes the intermediary
species M 9.46 kcal/mol more stable than the potential energy
of the sum of 1 and DBU.
What’s more, four transition statess(S,2S,3R)-syn-TS,
(S,2R,3S)-syn-TS, (S,2S,3S)-anti-TS, and (S,2R,3R)-anti-
TSswere obtained (Figure 1), and the frequency calculations
showed that all of them were true transition states with only
one imaginary frequency corresponding to the vibration for
the formation of the carbon-carbon bond. It was found that
anti-TSs are less favorable than syn-TSs, due to the large
steric repulsion between tert-butyl of imine and Ni-enolate;
the re facial attacking (S,2S,3R)-syn-TS is 1.16 kcal/mol of
lower less energy than the si facial alternative (S,2R,3S)-
syn-TS, while (S,2R,3R)-anti-TS is 1.97 kcal/mol more stable
Finally, the computed energy profile for the proposed
mechanism is presented in Figure 2. It can be seen that
pathways of syn-TSs are more favorable than those of anti-
TSs, with lower activation energies. Furthermore, the transi-
tion state (S,2R,3S)-syn-TS is structurally less favorable than
the alternative (S,2S,3R)-syn-TS, due to steric hindrance; thus,
(S,2S,3R)-syn-TS is found to be 1.16 kcal/mol more stable
than (S,2R,3S)-syn-TS, and correction between them for the
solvent effect with polarizable continuum models provides
a small difference in the activation of free energies of 0.94
kcal/mol in favor of the (S,2S,3R)-syn-TS structure. In
addition, the computations reveal that the (S,2S,3R)-syn-3a
structure produces 1.84 kcal/mol less energy than (S,2R,3S)-
syn-3a in the gas phase; when the solvent effect is considered,
there is also a small decrease in free energy to -34.74 kcal/
mol. Apparently, the results indicate that the reaction to
produce (S,2S,3R)-syn-3a is the most favorable process
among these four pathways when both thermodynamic and
kinetic factors are considered. The activation energy of this
reaction is calculated to be approximately 22.6 kcal/mol in
J. Org. Chem. Vol. 73, No. 21, 2008 8567

Wang et al.
FIGURE 1. Optimized geometries and their relative energies of four transition states.
TABLE 4. Calculated Total Energies, Thermal Correction Energies, Free Energies, and Their Relative Energies to the Same Reactant (Etot
,
Ecorr, Gtot, ∆Etot, ∆Ecorr, and ∆Gtot), and Relative Solvent-Corrected Free Energies (∆Gsolv
)
b
b
c
Ea (au)
Ecorr (au)
Ga (au)
∆Ea (kcal/mol) ∆Ecorr (kcal/mol) ∆Ga (kcal/mol) ∆Gsolv (kcal/mol)
R
-3598.992664 -3598.259356 -3598.298111
-3598.956646 -3598.224889 -3598.255385
-3599.046005 -3598.310144 -3598.341348
-3598.954805 -3598.222807 -3598.253753
-3599.043076 -3598.306964 -3598.338238
-3598.951307 -3598.218876 -3598.246989
-3599.039758 -3598.303805 -3598.334977
0.00
22.60
-33.47
23.76
-31.63
25.95
-29.55
23.98
0.00
21.61
-31.83
22.92
-29.83
25.39
-27.85
23.37
0.00
26.81
-27.13
27.84
-25.18
32.08
-23.13
30.38
0.00
24.22
-34.74
25.16
-32.78
30.05
-29.21
26.61
(S,2S,3R)-syn-TS
(S,2S,3R)-syn-3a
(S,2R,3S)-syn-TS
(S,2R,3S)-syn-3a
(S,2S,3S)-anti-TS
(S,2S,3S)-anti-3a
(S,2R,3R)-anti-TS -3598.954448 -3598.222098 -3598.249700
(S,2R,3R)-anti-3a -3599.037109 -3598.301061 -3598.332149
-27.89
-26.13
-21.36
-28.87
a Total energies and free energies of optimized structures were calculated by the HF/6-31G(d) method. b Thermal correction energies (Ecorr, 298 K)
were computed from the results of frequency calculations by the HF/6-31G(d) method with scale factor of 0.9135. c Solvent-corrected relative activation
free energies (∆Gsolv) were calculated with electrostatic contributions according to the polarizable continuum (PCM) models, with ꢀ ) 20.7 to mimic the
CH3COCH3 medium.
the gas phase and 24.2 kcal/mol in acetone, which are in
good accord with experimental results.
be 97:3, which is in excellent agreement with the experimental
enantioselectivities (>99%).
With the solvent-corrected computational activation energies
for the Mannich reaction, it was important to compute the de
values of the syn adducts from the aforementioned theoretical
activation energies and compare them with experimental data.
Since the Mannich reactions are subject to the Curtin-Hammett
principle, the (S,2S,3R)-3a/(S,2R,3S)-3a product ratio can,
therefore, be estimated by eq 1 and used to infer the ratio of
(2S,3R)-4a/(2R,3S)-4a, in which G(S,2S,3R) and G(S,2R,3S) are the
free energies that lead to the (S,2S,3R)-3a and (S,2R,3S)-3a
products. Moreover, ∆Gsolv is defined as ∆Gsolv ) ∆G +
δAGsolv, where δAGsolv is the relative free energy difference of
G(S,2S,3R)-solv and G(S,2R,3S)-solv with the PCM model. The relative
free energies Gsolv for the (S,2S,3R)-syn-TS and (S,2R,3S)-syn-
TS transition states are presented in Table 4. According to eq
1 and the solvent-corrected data of ∆Gsolv, the (S,2S,3R)-3a/
(S,2R,3S)-3a ratio for the Mannich syn adducts is calculated to
exp(-∆G(S,2S,3R) ⁄ RT)
[(S,2R,3S-3a] exp(-∆G(S,2R,3S) ⁄ RT)
G(S,2R,3S) ⁄ RT)] ) exp[-(∆Gsolv ⁄ RT)] (1)
[(S,2S,3R-3a]
)
) exp[-(G(S,2S,3R) -
Conclusions
In conclusion, we have developed a practical and highly
efficient enantio- and diastereoselective route to syn-configured
R,ꢀ-diamino acids using a direct asymmetric Mannich reaction.
A broad range of aryl-, heteroaryl-, and alkyl-derived imines
can all be employed under operationally simple and safe
conditions. The absolute configuration of one of the products
was determined, and a model for the intermediate was proposed.
Further studies will focus on mechanistic aspects such as
compatibility of aliphatic imines and further applications of other
8568 J. Org. Chem. Vol. 73, No. 21, 2008

Enantio- and DiastereoselectiVe Mannich Reactions
FIGURE 2. Energy profiles of pathways of (S)-Ni(II) complex Mannich reaction with syn-TSs and anti-TSs. The potential energy of the reactant
(R) is set to zero. The most favorable reaction pathway to (S,2S,3R)-syn-3a adducts is shown in blue, and the others are in red. Activation energies
are shown in italics, and in parentheses are the calculated solvent-corrected activation free energies with the SCRF method based on the polarizable
continuum models (PCM) with ꢀ ) 20.7 to mimic the CH3COCH3 medium.
(m 2H), 4.26-4.11 (m, 2H), 3.41 (d, J ) 12.9 Hz, 1H), 3.23 (t, J
) 17.7 Hz, 1H), 2.87-2.84 (m, 1H), 2.21-2.16 (m, 2H), 1.92-1.86
(m, 2H) ppm; 13C NMR (75 MHz, CDCl3) δ 180.3, 177.7, 173.2,
154.5, 143.3, 138.0, 136.5, 133.9, 133.7, 133.1, 132.9, 131.4, 130.1,
129.6, 129.1, 128.9, 128.8, 128.7, 128.5, 128.3, 127.8, 127.2, 126.4,
125.6, 123.2, 120.6, 77.3, 77.0, 76.7, 72.6, 70.2, 66.4, 63.5, 57.2,
55.3, 30.5, 22.9 ppm; MS (ESI, m/z) 737 [M + H]+, 759 [M +
Na]+; HRMS (ESI) calcd for C42H38N4NiO5 [M + Na]+ 759.2093,
found 759.2089.
chiral Ni(II) complexes in important carbon-carbon bond-
forming reactions. The results thereof will be reported in due
course.
Experimental Section
General Procedures for the Synthesis of (S,2S,3R)-3a. The
Ni(II) complex of glycine 1 (100 mg, 0.201 mmol) was dissolved
in acetone (10 mL). The R-amino sulfone derived from benzalde-
hyde 2a (73 mg, 0.211 mmol) and DBU (37 µL, 0.241 mmol) were
added under ambient conditions. The reaction mixture was then
stirred at room temperature for 4 h. The crude reaction mixture
was concentrated and then washed three times with water and brine,
and the combined aqueous phase was extracted with dichlo-
romethane (three times). The combined organic layers were dried
with Na2SO4, concentrated, and purified by flash column chroma-
tography (petroleum ether/ethyl acetate) to give (S,2S,3R)-3a as a
red solid.
General Procedures for the Synthesis of (2S,3R)-4a·HCl: The
crystallized complex (S,2S,3R)-3a (1 g, 1.4 mmol) was decomposed
by refluxing a suspension in a mixture of aqueous 6 N HCl (1 mL)
and MeOH (15 mL) for 30 min, until the red color of the solution
disappeared, as described previously. The reaction was cooled to
room temperature and then evaporated to dryness. Water (20 mL)
was added to the residue to form a clear solution, and this solution
was then separated by column chromatography on C18-reversed
phase (230-400 mesh) silica gel. Pure water as an eluent was
employed to remove the green NiCl2 and excess HCl; MeOH/water
(1/1) was then used to obtain optically pure product (2S,3R)-4a·HCl
Ni(II)-(S)-BPB/(2S,3R)-2-Amino-3-tert-butoxycarbonylamino-
3-phenylpropanoic Acid Schiff Base Complex 3a: Obtained as a
red solid by flash column chromatography (petroleum ether/ethyl
acetate 1:1), yield 93%; mp 242-243 °C; [R]22D ) +2154 (c 0.35
(246 mg, 96%): [R]22 ) +37 (c 0.38 g/100 mL, 6 N HCl, lit.17
D
[R]26 ) +39, c ) 0.52, H2O). The ligand BPB that decomposed
D
1
from (S,2S,3R)-3a was recovered by MeOH eluent (525 mg, 96%),
and the column chromatography was washed with 100 mL of
MeOH for further use.
g/100 mL, CHCl3); H NMR (300 MHz, CDCl3) δ 8.33 (d, J )
9.0 Hz, 1H), 7.97 (d, J ) 6.6 Hz, 2H), 7.62-7.51 (m, 6H),
7.36-7.13 (m, 8H), 6.78-6.70 (m, 2H), 6.45 (d, J ) 8.7 Hz, 1H),
4.71-4.69 (m, 1H), 4.23-4.18 (m 2H), 3.43 (t, J ) 12.9 Hz, 1H),
3.24-3.21 (m, 1H), 2.85-2.82 (m, 1H), 2.22-2.20 (m, 2H),
1.92-1.80 (m, 2H), 1.27 (s, 9H) ppm; 13C NMR (75 MHz, CDCl3)
δ 180.2, 177.7, 173.0, 153.8, 143.2, 138.3, 133.8, 133.7, 133.1,
132.8, 131.4, 129.8, 129.7, 128.9, 128.8, 128.7, 128.6, 128.3, 128.2,
128.0, 126.3, 125.6, 123.2, 120.5, 79.2, 77.3, 77.0, 76.7, 73.1, 70.1,
63.4, 57.1, 55.2, 30.4, 28.1, 22.8 ppm; IR (KBr) 704, 1163, 1252,
1493, 1635, 1664 (CdN), 1716 (CdO), 2976, 3396 cm-1; MS (EI,
m/z) 702 [M]+; HRMS (EI) calcd for C39H40N4NiO5 [M]+ 702.2352,
found 702.2358.
(2S,3R)-2,3-Diamino-3-phenylpropanoic Acid 4a·HCl: Ob-
tained as a white solid by column chromatography (MeOH/water
1:1), yield 96%; mp 187-189 °C; [R]22 ) +37 (c 0.38 g/100
D
1
mL, 6 N HCl); H NMR (300 MHz, D2O) δ 7.58-7.48 (m, 5H),
4.52 (d, J ) 10.2 Hz, 1H), 3.66 (d, J ) 10.2 Hz, 1H) ppm; 13C
NMR (75 MHz, D2O) δ 171.9, 133.6, 132.5, 132.1, 130.5, 56.3,
56.0 ppm; MS (ESI, m/z) 181 [M + H]+; HRMS (ESI) calcd for
C9H12N2O2 [M + Na]+ 203.0796, found 203.0797.
General Procedures for the Synthesis of (2S,3R)-6a·HCl: The
crystallized complex (S,2S,3R)-5a (1 g, 1.3 mmol) was decomposed
by refluxing a suspension in a mixture of aqueous 6 N HCl (1 mL)
and MeOH (15 mL) for 30 min, until the red color of the solution
disappeared, as described previously. The reaction was cooled to
room temperature and then evaporated to dryness. Water (20 mL)
Ni(II)-(S)-BPB/(2S,3R)-2-Amino-3-benzylcarbonylamino-3-
phenylpropanoic Acid Schiff Base Complex 5a: Obtained as a
red solid by flash column chromatography (petroleum ether/ethyl
acetate 1:1), yield 93%; mp 118-120 °C; [R]22D ) +1996 (c 0.53
g/100 mL, CHCl3); 1H NMR (300 MHz, CDCl3) δ 8.31(d, J ) 8.1
Hz, 1H), 7.97 (d, J ) 7.8 Hz, 2H), 7.60-7.52 (m, 6H), 7.37-7.13
(m, 13H), 6.86-6.68 (m, 3H), 4.96 (d, J ) 12.6 Hz, 1H), 4.87-4.70
(17) Zhang, H.-H.; Hu, X.-Q.; Wang, X.; Luo, Y. C.; Xu, P.-F. J. Org. Chem.
2008, 73, 3634.
J. Org. Chem. Vol. 73, No. 21, 2008 8569

Wang et al.
was added to the residue to form a clear solution, and this solution
was then separated by column chromatography on C18-reversed
phase (230-400 mesh) silica gel. Pure water as an eluent was
employed to remove the green NiCl2 and excess HCl; MeOH/water
(1/1) was then used to obtain optically pure product (2S,3R)-6a ·HCl
(397 mg, 96%). The ligand BPB that decomposed from (S,2S,3R)-
5a was recovered by MeOH eluent (525 mg, 96%), and the column
chromatography was washed with 100 mL of MeOH for further
use.
Na]+; HRMS (ESI) calcd for C17H18N2O4 [M + H]+ 315.1345,
found 315.1341.
Acknowledgment. We gratefully acknowledge financial
support from the State Key Program of Basic Research of China
(Grant 2006BAI01B02), the National Natural Science Founda-
tion of China (Grants 20721003 and 20872153), the Basic
Research Project for Talent Research Group from the Shanghai
Science and Technology Commission, and the 863 Hi-Tech
Program of China (Grants 2006AA020602).
(2S,3R)-Nꢀ-Cbz-2,3-Diamino-3-phenylpropanoic Acid 6a·HCl:
Obtained as a white solid by column chromatography (MeOH/water
1:1), yield 96%; mp 199-201 °C; [R]22 ) +48 (c 0.32 g/100
D
Supporting Information Available: Experiment procedures
and analytical and spectral characterization data for all compounds,
and crystallographic information files (CIF). This material is
available free of charge via the Internet at http://pubs.acs.org.
mL, 6 N HCl); 1H NMR (300 MHz, D2O) δ 7.47-7.35 (m, 10H),
5.14-5.13 (d, J ) 4.8 Hz, 2H), 3.72-3.70 (d, J ) 6.3 Hz, 1H),
3.64 (s, 1H) ppm; 13C NMR (75 MHz, D2O) δ 166.1, 154.0, 132.9,
132.1, 127.5, 127.2, 126.3, 126.1, 125.8, 125.4, 125.0, 124.5, 123.8,
64.2, 54.2, 51.7 ppm; MS (ESI, m/z) 315 [M + H]+, 327 [M +
JO8019169
8570 J. Org. Chem. Vol. 73, No. 21, 2008