Binaphthol-derived bisphosphoric acids serve as efficient organocatalysts for highly enantioselective 1,3-dipolar cycloaddition of azomethine ylides to electron-deficient olefins

Article abstract of DOI:10.1021/ja204218h

A variety of chiral bisphosphoric acids derived from binaphthols have been evaluated for enantioselective 1,3-dipolar cycloaddition reactions, revealing that the feature of the linker in the catalysts exerted great impact on the stereoselectivity. Among t

Full text of DOI:10.1021/ja204218h

ARTICLE
pubs.acs.org/JACS
Binaphthol-Derived Bisphosphoric Acids Serve as Efficient
Organocatalysts for Highly Enantioselective 1,3-Dipolar Cycloaddition
of Azomethine Ylides to Electron-Deficient Olefins
Long He,†,‡ Xiao-Hua Chen,† De-Nan Wang,† Shi-Wei Luo,*,† Wen-Quan Zhang,† Jie Yu,† Lei Ren,† and
Liu-Zhu Gong*,†
†Hefei National Laboratory for Physical Sciences at the Microscale and Department of Chemistry,
University of Science and Technology of China, Hefei, 230026, China
‡College of Chemistry and Chemical Engineering, China West Normal University, Nanchong, 637002, China
S Supporting Information
b
ABSTRACT: A variety of chiral bisphosphoric acids derived
from binaphthols have been evaluated for enantioselective 1,3-
dipolar cycloaddition reactions, revealing that the feature of the
linker in the catalysts exerted great impact on the stereoselec-
tivity. Among them, the oxygen-linked bisphosphoric acid 1a
provided the highest level of stereoselectivity for the 1,3-dipolar
cycloaddition reaction tolerating a wide range of substrates
including azomethine ylides, generated in situ from a broad
scope of aldehydes and R-amino esters, and various electron-
deficient dipolarophiles such as maleates, fumarates, vinyl ketones,
and esters. This reaction actually represents one of the most
enantioselective catalytic approaches to access structurally
diverse pyrrolidines with excellent optical purity. Theoretical calculations with DFT method on the formation of azomethine
ylides and on the transition states of the 1,3-dipolar cycloaddition step showed that the dipole and dipolarophile were
simultaneously activated by the bifunctional chiral bisphosphoric acids through the formation of hydrogen bonds. The effect of
the bisphosphoric acids on reactivity and stereochemistry of the three-component 1,3-dipolar cycloaddition reaction was also
theoretically rationalized. The bisphosphoric acid catalyst 1a may take on a half-moon shape with the two phosphoric acid groups
forming two intramolecular hydrogen bonds. In the case of maleates, one phosphate acts as a base to activate the 1,3-dipole, and
simultaneously, the two hydroxyl groups in the catalyst 1a may respectively form two hydrogen bonds with the two ester groups of
maleate to make it more electronically deficient as a much stronger dipolarophile to participate in a concerted 1,3-dipolar
cycloaddition with azomethine ylide. However, in the cases involving acrylate and fumarate dipolarophiles, only one hydroxyl group
forms a hydrogen bond with the ester functional group to lower the LUMO of the CꢀC double bond and another one is remained to
adjust the acidity and basicity of two phosphoric acids to activate the dipole and dipolarophile more effectively.
’ INTRODUCTION
motifs.3 Auxiliary-controlled 1,3-dipolar cycloaddition has been
a reliable method to produce optically pure pyrrolidine structural
motifs by using either the optically pure azomethine ylides or
dipolarophiles.4 Allway and Grigg first demonstrated that stoi-
chiometric amounts of chiral cobalt, manganese, and silver
complexes were able to promote the 1,3-dipolar cycloaddition
reactions of azomethine ylides derived from arylidene imines of
glycine with high enantioselectivity.5 The first catalytic enantio-
selective 1,3-dipolar cycloaddition reaction of azomethine ylides
was established independently by Zhang6a and Jørgensen6b who
respectively exploited chiral silver and copper catalysts to accel-
erate the reaction and control the stereochemistry. After these
two events, tremendous efforts have been directed toward the
design of new catalysts or ligands applicable to controlling
The rapid and facile construction of structurally diverse
heterocycles in enantiomerically pure form is among the most
important tasks in medically relevant fields, particularly in
chemical biology.1 Five-membered nitrogenous heterocycles, in
particular, pyrrolidines actually occupy a privileged position in
organic chemistry, because they are key structural motifs pre-
valent in numerous natural alkaloids as exemplified by the
molecules shown in Figure 1, and in biologically active unnatural
substances. Moreover, they also served as the chiral building
blocks or intermediates in the synthesis of natural alkaloids and
medicinally relevant compounds.2 Consequently, the develop-
ment of efficient methods to access optically pure pyrrolidine
skeletons are undoubtedly appealing in the organic synthesis.
The 1,3-dipolar cycloaddition reaction of azomethine ylides
with electronically poor olefins represents one of the most
efficient synthetic methods to access pyrolidine structural
Received: May 7, 2011
Published: July 22, 2011
r
2011 American Chemical Society
13504
dx.doi.org/10.1021/ja204218h J. Am. Chem. Soc. 2011, 133, 13504–13518
|

Journal of the American Chemical Society
ARTICLE
Figure 1. Naturally occurring alkaloids contain pyrrolidine motifs.
stereoselectivity in 1,3-dipolar cycloaddition reactions involving
azomethine ylides. As a result, many privileged bidentate ligands
turned out to give high levels of enantioselectivity over the last
decades.7 However, Lewis acid-catalyzed 1,3-dipolar cycloaddition
reactions are mostly sensitive to the substrates and usually one
chiral ligand or catalyst is highly enantioselective for one specific
type of dipolarophiles. Therefore, structurally diverse chiral ligands
are basically required to realize highly stereoselective 1,3-dipolar
cycloaddition reactions for different types of dipolarophiles, and
thereby somewhat hindering the practical use. In addition, a
catalytic amount of external base is used for the generation of
azomethine ylides. In recent years, organocatalytic 1,3-dipolar
cycloaddition reactions of azomethine ylides to dipolarophiles have
been discovered. R-Amino acids have found their applications to
the catalytic 1,3-dipolar cycloaddition of azomethine ylides to
electron-deficient olefins, but with a very low enantioselectivity
(up to 12% ee).8 The enones and enals activated by lowering the
LUMO through the formation of iminium intermediates with
chiral secondary amine catalysts proved to be highly reactive
toward the azomethine ylides in high stereoselectivity. However,
this process only works for the substrates capable of forming
unsaturated iminium species with amine andtherefore is principally
limited to enal and enone dipolarophiles.9 On the other hand, both
nitroolefins and imino esters could be simultaneously activated by
bifunctional thiourea catalysts to undergo a formal [3 + 2]
cycloaddition with stereocontrol.10 However, only nitroolefins
have been reported to be reactive substrates in this reaction.
Yamamoto has demonstrated that chiral N-triflyl phosphoramide
is used for an asymmetric 1,3-dipolar cycloaddition of diaryl
nitrones to ethyl vinyl ether to give the endo products in high
ee’s (up to 93% ee).11 Herein, we will present a three-component
1,3-dipolar cycloaddition of azomethine ylides with electron-defi-
cient olefins, which provided high levels of enantioselectivity for a
wide range of substrates including azomethines, generated in situ
from a wide scope of aldehydes and R-amino esters, and different
types of electron-deficient dipolarophiles such as maleates, fuma-
rates, vinyl ketones, and acrylates. In addition, a detailed theoretical
investigation on the reaction mechanism will also be reported.12
to be privileged bifunctional catalysts broadly applicable to
afford a large number of highly enantioselective organic
transformations.13 Initially, we envisioned that the bifunctional
phosphoric acid is theoretically able to form a chiral azomethine
ylide dipole IIa or IIb with an azomethine compound through
double hydrogen bonds. These chiral azomethine ylides were
presumably reactive toward electron deficient olefins to undergo
the 1,3-dipolar cycloaddition reaction in enantioselective manner
(eq 1).
Following up this consideration, we carried out the three-
component 1,3-dipolar cycloaddition reaction of 4-nitrobenzal-
dehyde (2a) with diethyl amino malonate (3a) and dimethyl
maleate (4a) in the presence of a chiral phosphoric acid and found
that the binol-derived phosphoric acids are able to catalyze the
reaction and a bisphosphoric acid 1a was identified as the catalyst
of choice, conferring exquisite levels of stereoselectivity for the
reaction capable of accommodating a wide scope of aldehydes
(eq 2).12 Subsequent studies found that the bisphosphoric acid
was also amenable to the 1,3-dipolar cycloaddition of 2,3-
allenoates with stereochemical control, yielding 3-methylenepyr-
rolidine derivatives with excellent enantioselectivity.14 Addition-
ally, on the basis of this general strategy, we established a
three-component 1,3-dipolar cycloaddition reaction of methyle-
neindolinones with azomethine ylides catalyzed by a 3,30-di-
(2-naphthyl) binol-based phosphoric acid, offering a facile entry
to spiro [pyrrolidin-3,30-oxindole] derivatives with high enan-
tioselectivity and structural diversity.15 Theoretical calculations
on this reaction suggested that both the azomethine ylide and the
methyleneindolinone are simultaneously activated by hydrogen-
bonds with the phosphoryl oxygen and hydroxyl of the
’ INITIAL CONSIDERATION OF THE BRØNSTED
ACID-CATALYZED THREE-COMPONENT REACTION
AND SUMMARY OF PRELIMINARY RESULTS
Binol-derived phosphoric acids containing a strong acidic
hydroxyl and a Lewis basic phosphoryl oxygen have been proven
13505
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
Figure 2. Bisphosphoric acids evaluated in this study.
phosphoric acid, respectively, to afford a stereocontrolled 1,3-
dipolar cycloaddition.15 Despite these findings, however, some
issues related with the 1,3-dipolar cycloaddition reaction cata-
lyzed by the bisphosphoric acid remain including that (1) the
effect of the feature of the linker in bisphosphoric acids on the
reaction should be clarified; (2) detailed mechanistic studies to
understand why this bisphosphoric acid is more stereoselective
than mono ones have not been conducted; and (3) the generality
for the dipolarophiles other than maleates has not been explored.
Thus, this article will be focused on detailed studies on the
reaction catalyzed by the bisphosphoric acids (1aꢀ1g) bearing
different linkers to addressthe issues mentioned above (Figure 2).
Table 1. Evaluation of the Bisphosphoric Acidsa
entry
1
yield (%)b
ee (%)c
1
2
3
4
5
6
7
8
1a
1b
1c
1d
1e
1f
96
33
79
27
74
26
50
89
98
92
45
10
23
76
59
96d
’ RESULTS AND DISCUSSION
The Preparation of Bisphosphoric Acids. The bisphospho-
ric acid catalysts 1aꢀ1g were prepared starting with BINOL,
which were first transformed into tetraols bearing various linkers re-
ferring to known reaction procedures established by Shibasaki16
and others,17 and followed by a reaction with 4 equiv of phos-
phorus oxychloride in the presence of pyridine and a subsequent
hydrolysis under acidic conditions. The synthetic details have
been shown in the Supporting Information.
1g
1a
a The reaction was performed in 0.1 mmol scale. b Isolated yield and a
c
single diastereomer was observed unless indicated otherwise. The ee
was determined by HPLC. d 1a was twice subjected to a short-path
column chromatography using extra pure silica gel (Merck) after
washing with HCl (2 M).
The 1,3-Dipolar Cycloaddition of 4-Nitrobenzaldehyde
with Diethyl Aminomalonate and Dimethyl Maleate Cata-
lyzed by Different Bisphosphoric Acids. Our previous evalua-
tion of binol-based phosphoric acids found that the
bisphosphoric acid 1a was the most effective and stereoselective
chiral catalyst for the three-component 1,3-dipolar cycloaddition
reaction of aldehydes with diethyl amino malonate and dimethyl
maleate, whereas the monophosphoric acids only delivered
moderate stereoselectivity.12 Moreover, the R-phenylglycine
methyl ester also participated in the reaction but at an elevated
temperature (50 °C). These preliminary results drew our atten-
tion to investigate the correlation between the structure of
bisphosphoric acids and reaction performance to seek the
optimal reaction conditions for the highly enantioselective 1,3-
dipoar cycloaddition reaction tolerating different substrates.
Thus, we reinvestigated the reaction of 4-nitrobenzaldhyde
(2a) with diethyl aminomalonate (3a) and dimethyl maleate
(4a) in dichloromethane at room temperature in the presence of
10 mol % of bisphosphoric acids 1aꢀg. Interestingly, the linker in
the bisphosphoric acids exerts great impact on the control of
stereoselectivity (Table 1). The replacement of the oxygen with
methylene led to a catalyst 1b, which provided a dramatically
reduced catalytic activity although the enantioselectivity eroded
slightly (entry 1 vs 2). The two carbon-bridged bisphosphoric acid
1c gave moderate yield with low enantioselectivity (entry 3). The
incorporation of heteroatoms other than oxygen into the linker
resulted in a poor enantioselectivity (entries 4ꢀ6). In particular,
the sulfonyl bridged bisphosphoric acid 1f showed poor catalytic
activity albeit with a moderate enantioselectivity (entry 6). The
introduction of phenyl substituents to the bisphosphoric acid, as
shown in 1g, led to a much lower catalytic efficacy and stereo-
controlling ability in comparison with 1a (entry 7 vs 1). The use
of acid washed 1a, which was twice subjected to a short-path
column chromatography using extra pure silica gel, gave a
comparable yield and enantioselectivity (entry 8). These results
in combination with the control reaction catalyzed by calcium
salt of 1a suggested that no metal contamination affected the
stereoselectivity.18
The 1,3-Dipolar Cycloaddition of 4-Nitrobenzaldehyde
with Diethyl Aminomalonate and Either 3-Butenone or
Methyl Acrylate Catalyzed by Bisphosphoric Acid 1a. Besides
the substrates examined in the preliminary report,12 additional
investigations were performed on the generality of the protocol
for vinyl ketones and acrylate esters (Table 2). Although acrylate
esters were previously investigated as dipolarophiles in the metal
complex-catalyzed asymmetric 1,3-dipolar cycloadditions, high
levels of enantioselectivity were only observed for the tert-butyl
acrylate7a,b,19 with exception of protocols reported by Kobayashi7f,g
and Wang.7h Moreover, 2-substituted acrylate esters, which
undergo the dipolar cycloaddition to give pyrrolidines bearing
an all-carbon stereogenic center, and methyl vinyl ketone (MVK)
have not been explored as dipolarophiles, yet. Delightedly, the
methyl vinyl ketone was well tolerated and participated in
smoothly phosphoric acid-catalyzed 1,3-dipolar cycloaddition
reactions with various aldehydes, producing the pyrrolidine
derivatives with high enantioselectivities ranging from 86% to
97% ee (entries 1ꢀ9). Interestingly, methyl acrylate also engaged
13506
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
Table 2. Scope of Aldehydes in the Asymmetric Three-
Component 1,3-Dipolar Cycloaddition Reaction Involving
Diethyl r-Aminomalonate with Vinyl Ketones and Estersa
Figure 3. X-ray structure of 7h .
entry
7
R1
R2
R3 time (h) yield (%)b ee (%)c
Table 3. Optimization of Conditions for the Reaction In-
volving r-Phenylglycine Methyl Estera
1
7a 4-NO2C6H4
7b 4-CNC6H4
CH3
CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
48
74
72
72
74
72
64
70
74
48
72
64
48
64
48
48
72
72
72
60
77
47
51
58
64
43
85
43
63
82
91
96
84
80
79
85
85
65
71
85
96d
96
87e
2
3
7c 4-MeO2CC6H4 CH3
4
7d 4-BrC6H4
7e 4-ClC6H4
7f 4-FC6H4
CH3
95
5
CH3
91
6
CH3
94
7
7g 2-NO2C6H4
7h 3-ClC6H4
7i 3-NO2C6H4
7j 4-NO2C6H4
7k 2-NO2C6H4
7l 4-CNC6H4
CH3
96d
86(96)f,g
97d
95
entry
1 (mol %)
solvent
temp. (°C)
yield (%)b
ee (%)c
8
CH3
9
CH3
1
2
3
4
1a (10)
1a (10)
1a (20)
1b (20)
CH2Cl2
CHCl3
CHCl3
CHCl3
RT
50
50
50
trace
55
ꢀ
97
97
98
10
11
12
OCH3
OCH3
OCH3
OCH3
OCH3
92
92
96
97e
78
13 7m 3-NO2C6H4
a The reaction was carried out in 0.2 mmol scale in a solvent (2 mL) with
3 Å MS (300 mg) for 48 h, and the ratio of 2a/8a/4a was 5/4/1.
b Isolated yield and a single diastereomer was observed unless indicated
otherwise. c The ee values were determined by HPLC analysis.
14
15
16
17
18
19
20
7n 4-BrC6H4
97
7o 4-MeO2CC6H4 OCH3
84
7p 4-NO2C6H4
7q 4-BrC6H4
7r 4-ClC6H4
7s 4-CNC6H4
7t 2-NO2C6H4
OCH3 Ph
OCH3 Ph
OCH3 Ph
OCH3 Ph
OCH3 Ph
94d
97d
94f
96d
98f
R-phenylglycine ester derivatives (Table 4). A wide scope of
aldehydes including aromatic, aliphatic, and unsaturated variants
were able to participate in the reaction. para-Substituted benzal-
dehydes underwent the 1,3-dipolar cycloaddition in excellent
yields and with high enantioselectivities ranging from 87 to 97%
ee (entries 1ꢀ6). meta-Nitro and cyanobenzaldehydes provided
96% and 92% ee’s, respectively, but a lower yield was obtained for
the cyanobenzaldehyde (entries 7 and 8). Interestingly, the
stereochemical outcomes are to some degree reliant on the
electronic properties of the ortho-substituents on benzaldehydes
and benefit from the electron-withdrawing group. For instance,
the highest ee value was found in the reaction of 2-nitrobenzal-
dehyde (entry 9) while comparably lower enantioselectivities
ranging from 83% to 91% ee were delivered to 2-haleganated
benzaldehydes (entries 10ꢀ12). Disubstituted benzaldehydes
were also suitable reaction components and participated in clean
reactions with 88ꢀ98% ee (entries 13ꢀ15). In contrast, aliphatic
aldehydes proved to be less reactive and gave much lower
enantiomeric excesses (entries 16 and 17). Cinnamaldehyde
derivatives were accommodated to give desired products in high
yields (96% and 89%, respectively) and with excellent stereo-
selectivities (96% and 98% ee, respectively, entries 18 and 19).
Although phenylpropiolaldehyde was a good reaction partner, a
moderate enantioselectivity was obtained (entry 20). Variation
of the ester pattern on either R-amino esters (8) or maleates (4)
led to very little change in the reaction conversion and the
stereoselectivity (entries 21 and 22). The configuration of 9e
(R1 = 4-BrC6H4) was assigned to be (2S,3S,4R,5S) by X-ray
analysis (Figure 4).
a The reaction was carried out in 0.1 mmol scale in dichloromethane
(1 mL) with 3 Å MS (150 mg), and the ratio of 2/3a/6 was 1.2/1/5.
b Isolated yield and a single diastereomer was observed unless indicated
otherwise. c The ee values were determined by HPLC analysis. d >99/1
dr. e 99/1 dr. f 98/2 dr. g After a single recrystallization from ethyl acetate.
in the reaction to give high yields and excellent enantioselec-
tivities (entries 10ꢀ15, 84ꢀ97% ee). More importantly, methyl
2-phenylacrylate underwent the three-component 1,3-dipolar
cycloaddition reaction with a variety of aldehydes and amino-
malonate to furnish pyrrolidine derivatives possessing an all-
carbon quaternary stereogenic center in good yields (65ꢀ85%)
and with excellent enantioselectivities (entries 16ꢀ20, 94ꢀ98%
ee).20 The configurations of the products were assigned to be
(4R, 5S) by X-ray analysis of 7 h (Figure 3).
The 1,3-Dipolar Cycloaddition of Aldehydes with Mal-
eates and Diethyl r-Amino Esters. However, the extension
ofthe reaction conditions to R-phenylglycinemethyl esterfailed to
undergo the 1,3-dipolar cycloaddition reaction (Table 3, entry 1).
Encouragingly, the reaction proceeded at elevated temperature
(50 °C) in chloroform to give the desired product in a high yield
and with an excellent enantioselectivity (entry 2).12 Interestingly,
all carbon-bridged bisphosphoric acid 1b was also able to offer
excellent enantioselectivity (entry 4). Interms of the efficiency and
stereochemical control, 1a is still the catalyst of choice. The
presence of 20 mol % of 1a ensured a clean reaction in 92% yield
and with 97% ee (entry 3).
Additionally, the generality for R-arylglycine methyl esters
other than methyl R-phenylglycine was explored (Table 5). The
substitution on the benzene ring was well tolerated to give a clean
With the optimal conditions in hand, we explored the
generality of the protocol for other different aldehydes and
13507
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
Table 4. Scope of Aldehydes in the Asymmetric Three-
Component 1,3-Dipolar Cycloaddition Reaction Involving
r-Phenylglycine Estersa
Table 5. Scope of r-Arylglycine Methyl Estersa
entry
10
Ar
time (h)
yield (%)b
ee (%)c
entry
9
R1
R2
R3
yield (%)b ee (%)c
1
2
3
4
5
6
7
8
9
10a
10b
10c
10d
10e
10f
4-ClC6H4
60
54
54
60
64
48
50
52
48
96
96
95
94
93
96
91
97
92
97
94
98
95
95
98
96
95
91
3-ClC6H4
1
9b 4-CF3C6H4
9c C6H5
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
CH3 CH3
98
90
96
93
96
86
86
65
98
96
95
96
94
87
84
58
61
96
89
83
95
95
94
89
87
97
94
94
96
92
96
87
83
91
96
88
98
75
74
96
98
78
92
95
2-ClC6H4
2
4-FC6H4
3
9d 4-CNC6H4
9e 4-BrC6H4
9f 4-ClC6H4
9g 4-FC6H4
9h 3-NO2C6H4
9i 3-CNC6H4
9j 2-NO2C6H4
9k 2-FC6H4
9l 2-ClC6H4
3-FC6H4
4
2-FC6H4
5
10g
10h
10i
4-MeOC6H4
4-MeO2CC6H4
3-Cl, 4-MeOC6H3
6
7
8
a The reaction was carried out in 0.2 mmol scale in chloroform (2 mL)
with 3 Å MS (300 mg), and the ratio of 2a/8/4a was 5/4/1. b Isolated
yield and a single diastereomer was observed unless indicated otherwise.
c The enantiomeric excesses were determined by HPLC analysis.
9
10
11
12 9m 2-BrC6H4
13
14
15
16
17
18
19
20
21
22
9n 2-Cl,4-FC6H3
9o 3-Cl,4-FC6H3
9p 4-Cl,3-FC6H3
9q c-C6H11
Table 6. Optimization of Reaction Conditions of the Asym-
metric Three-Component 1,3-Dipolar Cycloaddition Reac-
tions Involving Methyl Phenylalaninea
9r c-C3H5
9s 4-MeOC6H4CHdCH CH3 CH3
9t 4-NO2C6H4CHdCH CH3 CH3
9u PhCtC
CH3 CH3
CH3 C2H5
C2H5 CH3
9v 4-NO2C6H4
9w 4-NO2C6H4
a The reaction was carried out in 0.2 mmol scale in chloroform (2 mL)
with 3 Å MS (300 mg) for 48ꢀ72 h and the ratio of 2/8/4 was 5/4/1.
However, a messy reaction was observed for n-alkyl aldehydes and tert-
butyl aldehyde did not participate in the 1,3-dipolar cycloaddition.
b Isolated yield and a single diastereomer was observed unless indicated
otherwise. c The ee values were determined by HPLC analysis.
entry
temp (°C)
solvent
CHCl3
yield (%)b
ee (%)c
1
2
3
50
70
70
30
88
84
94
98
96
ClCH2CH2Cl
PhCH3
a The reaction was carried out in 0.2 mmol scale in solvent (2 mL) with 3
Å MS (300 mg) for 48 h, and the ratio of 2a/11a/4a was 5/4/1.
b Isolated yield and a single diastereomer was observed unless indicated
otherwise. c The enantiomeric excesses were determined by HPLC
analysis.
incomplete reaction albeit with a high enantioselectivity of 94%
ee, presumably due to the lower reactivity in comparison with the
R-aryl analogues (Table 6, entry 1). Thus, the reaction condi-
tions were reoptimized to improve the efficiency. Screening of
reaction parameters including solvents and temperature eluci-
dated that the best results in terms of both the yield and
stereoselectivity were achieved when the reaction was conducted
in 1,2-dichloroethane at 70 °C by using 20 mol % of 1a (entry 2).
The reoptimized conditions were then applied to structurally
different R-alkyl amino esters (Table 7). Among the methyl
phenylalanine derivatives, the substituent on the phenyl group
exerted very little impact on the stereoselectivity. The substitu-
tion of the phenylalanine with either electron-withdrawing or
donating group was tolerated with exquisite enantioselectivities
ranging from 97% to 98% ee (entries 1ꢀ8). Interestingly,
methyl R-aminobutyrate also participated in the 1,3-dipolar
Figure 4. The X-ray structure of 9e.
cycloaddition reaction with excellent enantioselectivities regard-
less of the electronic feature and the position of the substituents
(94ꢀ98% ee, entries 1ꢀ8). The reaction involving a disubsti-
tuted phenylglycine methyl ester occurred with a high stereo-
selectivity as exemplified by 3-chloro-4-methoxyphenyl glycine
methyl ester (91% ee, entry 9).
However, the further application of the optimized conditions
to the reaction involving racemic methyl phenylalanine led to an
13508
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
Table 7. Scope of Phenylalanine Esters in the Asymmetric
Three-Component 1,3-Dipolar Cycloaddition Reactionsa
Table 9. Generality for Amino Esters and Aldehydesa
entry 14
Ar
R
R0 yield (%)b endo/exoc ee (%)d
1
2
3
4
5
6
7
14b 4-BrC6H4
14c 4-ClC6H4
CO2Et Et
CO2Et Et
89
77
83
80
93
99
99
10/1
9/1
95
entry
12
R
time (h)
yield (%)b
ee (%)c
91(22)
96(47)
97(23)
92
97e (10)
50f (2)
14d 4-CNC6H4 CO2Et Et
14e 3-NO2C6H4 CO2Et Et
14f 2-NO2C6H4 CO2Et Et
8/1
1
2
3
4
5
6
7
8
9
12a
12b
12c
12d
12e
12f
Ph
60
48
48
48
48
50
42
46
72
88
82
88
92
92
81
65
73
79
98
96
97
98
98
98
97
98
94
9.5/1
8/1
4-ClC6H4
4-BrC6H4
2-FC6H4
14g 4-NO2C6H4 Ph
Et
3.8/1
14h 4-NO2C6H4 Bn
a The reaction was carried out in 0.1 mmol scale in toluene (1 mL) with 3
Me
4/1
4-MeOC6H4
4-CH3CO2C6H4
4-FC6H4
Å MS (150 mg) at 25 °C for 24ꢀ48 h, and the ratio of 2a/3/13 was 1.2/
1/5. b Isolated yield. Determined by H NMR. d The ee values were
determined by HPLC analysis and the ee in parentheses is for the minor
diastereomer. e at 50 °C. f At 70 °C.
c
1
12g
12h
12i
3-FC6H4
Me
a The reaction was carried out in 0.2 mmol scale in 1, 2-dichloroethane
(2 mL) with 3 Å MS (300 mg), and the ratio of 2a/11/4a was 5/4/1.
b Isolated yield and a single diastereomer was observed unless indicated
otherwise. c The ee values were determined by HPLC analysis.
Table 8. Optimization of Reaction Conditions for Dimethyl
Fumaratea
Figure 5. X-ray structure of 14c..
solvents found that chloroform and 1,2-dichloroethane could
deliver high enantioselectivities of 96% and 95% ee, respectively
(entries 2 and 3), whereas poor stereoselectivities were observed
in THF and acetonitrile (entries 4 and 5). Delightedly, the best
results in terms of reaction efficiency and stereoselectivity were
attained by conducting the reaction in toluene (entry 6).
With the optimal conditions for the 1,3-dipolar cycloaddition
of dimethyl fumarate in hand, we explored the generality for the
aldehydes and amino esters (Table 9). The substitution of benz-
aldehyde derivatives with electron-withdrawing group at either
the para- or meta-position was well operative in the 1,3-dipolar
cycloaddition with dimethyl fumarate (13), leading to highly
substituted pyrrolidine derivatives in high yields and with good
diastereomeric ratios and excellent ee values (entries 1ꢀ5). More
interestingly, the ethyl R-phenylglycine gave a quantitative yield
and excellent enantioselectivity, but the diastereomeric ratio
was unsatisfactory (entry 6). However, a moderate enantioselec-
tivity was obtained in the case involving methyl phenylalanine
(entry 7). The absolute configuration was assigned to (3R,4R,5S)
by X-ray analysis of 14c (Figure 5).
Computational Studies on the Reaction Mechanism by
DFT Calculation. The BINOL-derived phosphoric acid catalysts
acted simultaneously as an acid and a base to activate both the
nucleophile and electrophile in many reaction mechanisms.21
To gain deeper insight into this asymmetric 1,3-dipolar cyclo-
addition reaction involving BINOL-derived bisphosphoric acid
catalysts, full DFT calculations were performed. The B3LYP
functional22 combined with the 6-31G(d) basis set,23 as imple-
mented in the Gaussian03 program,24 was used for geometry
optimization of all intermediates and transition state structures.
entry
solvent
CH2Cl2
yield (%)b
endo/exoc
ee (%)d
1
2
3
4
5
6
92
71
82
38
78
82
2/1
73
96
95
10
40
97
CHCl3
7.5/1
8/1
ClCH2CH2Cl
THF
1/2
CH3CN
PhCH3
1/1
8.5/1
a The reaction was carried out in 0.1 mmol scale in solvent (1 mL) with 3
Å MS (150 mg) at 25 °C for 24 h, and the ratio of 2a/3a/13 was 1.2/1/
5. b Isolated yield. c Determined by H NMR. d The ee values were
1
determined by HPLC analysis.
cycloaddition in a good conversion and with excellent enantios-
electivity although a prolonged time was required (entry 9).
The 1,3-Dipolar Cycloaddition of Fumarate with Diethyl
Aminomalonate and Aldehydes. In comparison with maleates,
fumarates represent challenging dipolarophiles in the asym-
metric 1,3-dipolar cycloaddition reaction. Jørgensen and co-
workers accomplished a successful 1,3-dipolar cycloaddition
reaction of fumarates with azomethine ylides using a zinc com-
plex of tert-BOX ligand while the highest level of enantioselec-
tivity reached 90% ee.6b When we applied the optimal conditions
to the 1,3-dipolar cycloaddition involving dimethyl maleate to
dimethyl fumarate, the reaction proceeded smoothly to give the
cycloadduct in a high yield but with a moderate enantioselectivity
and a low diastereoselectivity (Table 8, entry 1). Screening of
13509
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
Figure 6. Calculated structures of catalyst, substrates and key intermediates with relative energies in enthalpy (blue) and Gibbs free energy (red) in
kcal/mol and distance in angstrom.
To save computational time, the naphthyl moieties in the catalyst
were replaced with phenyls and ethyl ester in the substrate with
methyl ester.
The optimized structures of the catalyst, cat-1/2, and the
substrates, amino-malonate and maleate, are shown in Figure 6.
The BINOL-derived bisphosphoric acid cat-1/2 take on a half-
moon shape with the two phosphoric acid groups forming two
intramolecular hydrogen bonds (H-bonds). The calculated
H-bond lengths are about 1.6ꢀ1.7 Å implying quite strong
H-bonding interactions. This indicated that the linkage between
the two BINOLs is proper for the two phosphoric acid groups to
Figure 7. Proposed models of the nucleophilic addition of amino-
malonate to benzadehyde.
interact with each other. One of the two H-bonds formed on the
inside of the half-moon shaped structure, and the other on the
outside. The outside one should be easier to open and accept
substrate to form the active complex by H-bonds. The inside one
may play a proton-shuttle role to adjust the acidity and/or
basicity of two phosphoric acid groups acting as a proton donor
or acceptor in the reaction process.
The ester groups in the amino-malonate could assume various
conformations. The most stable one, amino malonate, as shown
in Figure 6, have the carbonyl oxygen atoms pointing toward the
two hydrogen atoms of amine, respectively. The two ester groups
in maleate prefer to a perpendicular orientation (T-shape) to
avoid the repulsive interaction between the two carbonyl oxygen
atoms. These structural features will be relevant to catalysis. In
addition, there are two isomers for the key intermediate azo-
methine ylide, formed by condensation of amino-malonate with
benzaldehyde. The two bulky groups of phenyl ring and diester
methine may have trans- or cis- conformations. The trans-isomer
was predicted to be more stable than the cis-isomer by about
ca. ∼10 kcal/mol. Consequently, the less stable cis-isomer is not
considered in the following computations of 1,3-dipolar cyclo-
addition reactions.
The azomethine ylide is a key intermediate in this reaction.
Hence, the formation of azomethine ylide was investigated in
detail first. It was hypothesized that a BINOL-derived phospho-
ric acid catalyst can act simultaneously as an acid and a base to
promote the reaction between an electrophile and a nucleophile.
We thought the BINOL-derived bisphosphoric acid catalyst can
play the same roles with additional proton-shuttling movement
from the inside H-bond to adjust the acidity and basicity of
both phosphoric acid groups. On the basis of the optimized
13510
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
Scheme 1. Proposed Catalytic Cycle for the Formation of Azomethine Ylide
geometries of BINOL-derived bisphosphoric acids shown in
Figure 6, we proposed a model for the nucleophilic addition of
amino-malonate to benzaldehyde (Figure 7). According to this
model, the outside, half-moon shaped, H-bond in the BINOL-
derived bisphosphoric acid catalyst is opened when substrate
approaches to release the phosphoric OH acid and PdO base.
The phosphoric OH group acts as an acid to protonate the
benzaldehyde, and the PdO group as a base to accept the amino-
malonate. In both cases, H-bonding interaction is the intrinsic
driving force. The inside H-bond may significantly improve the
acidity and basicity simultaneously by proton shifting between
the two phosphoric acid groups. The two substrates, benzalde-
hyde and amino-malonate, may approach backward or forward
each other, respectively. The bulky groups of both substrates
point in opposite direction and away from the BINOL-frame-
work of the catalyst to reduce steric interactions. As such, there
can be two possible complexes in which the catalyst can activate
both reactants simultaneously to promote the nucleophilic
addition of amine to carbonyl carbon. A proposed catalytic cycle
for the formation of azomethine ylide is shown as Scheme 1.
The calculated results for the formation of azomethine ylide
are shown in Figure 8. The formation of either complex,
Complex-1-R or Complex-1-S, in Figure 8 was favorable with a
binding energy of about ꢀ16.6 (ΔH) or 5.4 (ΔG) kcal/mol for
Complex-1-S. The Complex-1-S was predicted to be more
preferable than Complex-1-R about 2.7 (ΔH) or 1.86 (ΔG)
kcal/mol. The subsequent location of transition state of the
CꢀN bond forming step gave a similar prediction that TS-1-S
was more stable than TS-1-R by about 2.7 (ΔH) or 1.9 (ΔG)
kcal/mol. Again, the calculated results of the intermediate amino
alcohol indicated also thatINT-1-S was more stable than INT-1-R
by about 2 kcal/mol. This predicted destabilization from Com-
plex-1-R to INT-1-R may come from the pseudo 1,3-diaxial
interaction between carbonyl oxygen atom and one ester group
of amino-malonate. The calculated activation barrier for the CꢀN
bond formingstep was ∼6.49(ΔG) kcal/mol. This somewhat low
activation barrier indicated that the catalyst is highly efficient and
the proton shifting between the two phosphoric acids likely
assisted in the nucleophilic addition process.
The next step is the dehydration of the amino alcohol
intermediate, INT-1-R/S. The dehydration process can also be
promoted by the BINOL-derived bisphosphoric acid catalyst.
Proton transfer from the protonated amine of INT-1-R/S to
protonated hydroxyl of INT-2-R/S gives the precursor of dehy-
dration. The located complex structures are shown as INT-2-R/S
in Figure 8. Such a perturbation is endergonic by ∼3 kcal/mol.
13511
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
Figure 8. Located structures of complexes, transition states for the procedure of formation of azomethine ylide. Labeled distance in angstrom and
relative energies in enthalpy (blue) and Gibbs free energy (red) in kcal/mol.
The located structure of intermediate INT-2-R/S indicated that
the breakage of CꢀO bond may also be assisted by one of the
ester groups by an intramolecular H-bonding interaction as
shown in INT-2-R/S. These two structures were predicted to
be of the same stability although the HꢀN was orientated in
opposite directions in these two structures. The CꢀO bond
breaking process can be promoted by the catalyst and the ester
group, simultaneously. The located TSs were shown as TS-2-R/S,
in which the hydroxyl leaving group is protonated by phosphoric
acid and concurrently stabilized by an H-bonding interaction
with one ester group. TS-2-S was predicted to be more favorable
than TS-2-R by ∼3.5 kcal/mol due to the preferred H-bonds
between HꢀN and phosphoryl oxygen in TS-2-S. The next
complex of catalyst and protonated imine (intermediate
13512
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
Figure 9. Energy profile for the formation of azomethine ylide.
Figure 10. Located stable structures of complexes of catalyst with dipole and dipolarphiles, and binding energies in enthalpy (blue) and Gibbs free
energy (red).
iminium) with H2O was identified as INT-3-R/S. INT-3-S was
predicted to be more stable than INT-3-R by about ∼7 kcal/mol
due to the lack of H-bonding interactions between HꢀN and
phosphoryl oxygen in INT-3-R.
azomethine ylide was produced. In the beginning of the
formation of azomethine ylide, the formed complex between
catalyst and azomethine ylide should take the orientation as
INT-5-R/S, which were similar as those in the TS. The more
stable complexes will be formed with a slight perturbation of the
azomethine ylide, shown as complex-cat-dipole-1/2/3 in Fig-
ure 10, which were predicted to be more stable than INT-5-R/S
by about 10 kcal/mol.
The energy profile for the formation of azomethine ylide from
the nucleophilic addition reaction of amino-malonate to benzal-
dehyde, catalyzed by BINOL-derived bisphosphoric acid, is
shown in Figure 9. The rate-determining step was predicted to
be the elimination of H2O, with an activation barrier of about
13 kcal/mol. It is lower compared to that in the formation of
imine catalyzed by proline25 or prolinamide.26 This indicated
bisphosphoric acid should be a more effective catalyst to promote
the formation of imine.
Release of the water molecule from the complex of INT-3-R/S
and followed by a slight perturbation of the protonated imine, the
precursor structure, for the formation of azomethine ylide were
ascribed as INT-4-R/S. Similarly, INT-4-R was less stable than
INT-4-S by about ∼9 kcal/mol. The two phosphatic groups
acted in concert now to promote the deprotonation of CꢀH to
form azomethine ylide, one H-bonded HꢀN to stabilize the TS
and the other one pointed to the HꢀC with one oxygen atom of
the phosphate as a base to deprotonate the H of the HꢀC bond.
The located TSs were shown as TS-3-R/S. Once again, TS-3-S
was predicted to be more favorable than TS-3-R by about ∼4
kcal/mol. After the shift of H of the HꢀC bond from iminum to
phosphate, the catalyst was regenerated and the key intermediate
13513
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
Figure 11. Proposed transition state models for the 1,3-dipolar cycloaddition reaction.
Figure 12. Located structures of complexes and TSs of 1,3-dipolar cycloaddition reaction of azomethine ylide with maleate.
The 1,3-dipolar cycloadditions of azomethine ylide with the
electron deficient CdC bond of maleate, fumarate, and acrylate
catalyzed by cat-2 were performed next, respectively. The high
enantioselectivity observed experimentally for the 1,3-dipolar
cycloaddition indicated a concerted process. The bisphosphoric
acids may function similarly as in the process of formation of
azomethine ylide. One phosphate acts as a base or H-bond
acceptor to activate the 1,3-dipole, another as an acid or H-bond
donor to make dipolarophile more electron deficient. The
structures of the complexes between cat-2 and 1,3-dipole of
azomethine ylide discussed previously as INT-5-R/S in Figure 8
and complex-cat-dipole-1/2/3, which correspond to other three
different orientations of the dipole of azomethine ylide in the
formed complex, are shown in Figure 10. The latter were
predicted to be more favorable than the former ones by about
10 kcal/mol due to two better H-bond interactions between cat-2
and the dipole of the azomethine ylide. In complex-cat-dipole-3,
the dipole and catalyst overlap. The si-face of the dipole was
blocked by the catalyst completely. The dipolarophile ap-
proaches the dipole without interesting interaction with the
catalyst. In complex-cat-dipole-1/2, the si-face of the dipole is
opened and easily accessible for approach by the dipolarophile.
In light of these structures, we located the complexes between
cat-2 and different dipolarophiles. For maleate, the twisted cis-
diester groups can approach bisphosphoric acid easily, and the
cat-2 may open its double H-bonds to adopt the maleate by
forming two new and stronger H-bonds, shown as complex-cat-
maleate, with a C2 axis, in Figure 10. The binding energy in
enthalpy was predicted to be ∼11 kcal/mol for complex-cat-
maleate and is nearly the same as that of complex-cat-dipole-1.
13514
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
Figure 13. Located structures of complexes and TSs of 1,3-dipolar cycloaddition reaction of azomethine ylide with acrylate.
These high binding enthalpies indicated that there are strong
interactions between the catalyst and both the dipole and the
dipolarophile. Thus, a concerted 1,3-dipolar cycloaddition pro-
cess could be promoted by the catalyst with both components
binding with the catalyst simultaneously. Similarly, both fumarate
and acrylate may form complexes with cat-2 by two H-bonds
between the carbonyl oxygen of the ester and two OH groups of
the catalyst as shown in Figure 10. However, the binding
enthalpy was predicted to be about 5 kcal/mol, less than half
of that of complex-cat-dipole and complex-cat-maleate.
In terms of the structural features of complex-cat-maleate and
complex-cat-dipole-1, we propose two possible ways for the cyclo-
addition process, Model-I and Model-II, as shown in Figure 11.
The structure of complex-cat-maleate has C2 symmetry and the
electron deficient dipolarphile maleate is highly activated by two
H-bonds. The approach of the 1,3-dipole to maleate from either
side is favorable. When the 1,3-dipole approaches maleate, the
basic oxygen of phosphate may form an H-bond with HꢀN of
the 1,3-dipole to direct its approaching and stabilizing the
coming TS of the cycloaddition process shown as Model-I. On
the basis of the structure of complex-cat-dipole-1, when maleate
approaches 1,3-dipole, the OH of the phosphate may form an
H-bond with the carbonyl oxygen of maleate to direct its
approach and activate the dipolarophile as shown as Model-II.
The re-facial approach of dipole to maleate was labeled as ‘-Ma’,
while the si-facial approach as ‘-Mi’.
the electron rich carbon of the 1,3-dipole than the electron deficient
one, with a difference of forming bond length of 0.33 Å in TS-
model-I-Ma and 0.14 Å in TS-model-I-Mi. The selectivity possibly
comesfromthedifferentinteractionsbetween the phosphate of cat-
2 and the ester group of dipole. As shown in the TSs, the basic
oxygen of the phosphate is pointing to its back. In TS-model-I-Ma,
the ester group of the dipole was located in the front, so the basic
oxygen of the ester was slightly away from the oxygen of the
phosphate and close to the positively charged phosphorus atom.
While in TS-model-I-Mi the ester group of the dipole was located in
the back, and the basic oxygen of the ester was close to the oxygen
of the phosphate and far away from the positively charged
phosphorus atom. The repulsive interaction between the two basic
oxygens and the attractive interaction between the basic oxygen and
positively charged phosphorus atom, that is, the attractive ion pair
that is interacting, made TS-model-I-Ma more favorable than
TS-model-I-Mi, resulting in the observed stereoselectivity.
From the located structure of complex-cat-dipole-1 shown in
Figure 10, the inside H-bond remains and the outside H-bond
opens to the dipole to form two new H-bonds. Because one side
of the dipole was blocked by the catalyst framework, the
dipolarophile maleate could approach only from the si-face of
the dipole to give the Model-II type cycloaddition. The located
TS structure was shown as TS-model-II-Mi in Figure 12, which is
the most stable one but not corresponding to the observed major
product. In TS-model-II-Mi, the dipole interacted with the
catalyst in the same way as in TS-model-I-Ma and TS-model-I-
Mi. However, there is only one H-bond between the catalyst and
maleate, so the maleate was less activated than that in TS-model-
I-Ma and TS-model-I-Mi, and the TS-model-II-Mi was predicted
to be less stable by about 2 kcal/mol than TS-model-I-Ma.
Substitution of one ester group of maleate with H atom gave
the acrylate with similar 1,3-dipolar cycloaddition patterns as
The located complexes andTS structures are shown in Figure 12.
Complex-model-I-Ma is more favorable than complex-model-I-Mi
by about 2 kcal/mol. TS-model-I-Ma, which corresponds to the
major product observed experimentally, was predicted to be more
stable than TS-model-I-Mi by about 4 kcal/mol. The forming
CꢀC bond distances were predicted to be 2.1ꢀ2.4 Å, indicating a
concerted cycloaddition with a slight earlier nucleophilic attack of
13515
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
Figure 14. Located structures of complexes and TSs of 1,3-dipolar cycloaddition reaction of azomethine ylide with fumarate.
Model-I and Model-II in Figure 11. The located complex and TS
structures were shown in Figure 13. Complex-model-II-aMa/
aMi were predicted to be more favorable than complex-model-I-
aMa by about 2 kcal/mol. The stability of the corresponding TSs
was predicted with the same tendency as that of the complexes.
TS-model-II-aMa, which afforded the observed major product,
was more stable than TS-model-I-aMa by about 1 kcal/mol. This
meant Model-II is slightly better than Model-I for the 1,3-
cycloaddition of azomethine ylide with acrylate catalyzed by
BINOL-derived bisphosphoric acid. In both TS-model-II-aMa
and TS-model-I-aMa, the nucleophilic attack of the 1,3-dipole on
the β-carbon of acrylate ocurs quite early with CꢀC distance of
∼2.07 Å, compared to the other forming CꢀC bond distance of
∼2.8 Å. Thus, the 1,3-dipolar cycloaddition reaction seemingly
undergoes the Michael addition first, followed by an annulation
to furnish the cycloadduct. However, the structure of the
intermediate generated from the Michael addition could not be
located, and therefore, the reaction may still proceed via a con-
certed rather than a sequential pathway. If the electron deficient
carbon of the 1,3-dipole as nucleophile is added to R-carbon of
acrylate, as in TS-model-II-aMi, the resulting TS has a higher
activation barrier. TS-model-II-aMi was predicted to be less
stable than TS-model-II-aMa by about 5.6 kcal/mol. In case of
fumarate as the dipolarophile, similar calculated results were
obtained. The located complex and TS structures were shown in
Figure 14.
However, for acrylate and fumarate, the ester groups may form
one or two H-bonds with the catalyst as Model-I and Model-II to
activate the R,β-CꢀC double bond. The calculated results sug-
gest that the 1,3-dipolar cycloaddition reaction may proceed via a
concerted rather than a sequential pathway although the attack
on the β-carbon occurs quite earlier than that on the R-carbon of
the activated ester by the nucleophilic center of the 1,3-dipole in
both Model-I and Model-II, considering that the structure of the
intermediate generated from the attack on the β-carbon could
not be located. Moreover, the Model-II with one H-bond to the
catalyst was slightly better than Model-I with two H-bonds for
the nucleophilic attack on the β-carbon, implying that the one
H-bonding interaction with the catalyst is sufficient to activate
the acrylate and fumarate dipolarophiles, while the H-bond
remaining in the catalyst may adjust the acidity and basicity of
the two phosphate groups, simultaneously, by H-shifting be-
tween the two phosphate groups to activate the acceptor and
donor appropriately. This BINOL-derived bisphosphoric acid
may present a different activating model for substrates with
various functional groups to facilitate different types of reaction
with selectivity.
’ CONCLUSION
In summary, we have developed highly enantioselective 1,3-
dipolar cycloaddition reactions using chiral bisphosphoric acids
derived from binaphthols as catalysts. The linkage in bispho-
sphoric acids on the reaction were clarified as incorporation of an
oxygen atom into the linker resulted in the best reactivity and
enantioselectivity. This organocatalytic 1,3-dipolar cycloaddition
reaction by catalyst 1a is applicable to a wide range of azomethine
ylides derived from various aldehydes and R-amino esters with
electron-deficient dipolarophiles, providing a highly enantio-
selective method to access structurally diverse pyrrolidines with
These computed results indicated that activating patterns in
these reactions were different from maleate with two ester groups
with a cis-configuration and for acrylate and fumarate with only
one ester group in the dipolarophile. The two OH groups of two
phosphoric acids in the catalyst may form two H-bonds with both
ester groups of maleate, respectively, to make it more electron
deficient and a much stronger dipolarophile to conduct a
concerted 1,3-dipolar cycloaddition with azomethine ylide.
13516
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
excellent optical purity. Theoretical calculations with the DFT
method indicated that the bisphosphoric acids catalyst 1a may
take on a half-moon shape with the two phosphoric acid groups
forming two intramolecular hydrogen bonds. It may take two
models to promote reaction: two H-bonds are opened to activate
electrophile by its two hydroxyls with two H-bonding interac-
tions simultaneously or only one H-bond is opened and the an-
other one remains to play a proton-shuttle role to adjust the
acidity and/or basicity of two phosphoric acid groups acting as a
proton donor or acceptor in the reaction process. Computation
on the formation of key intermediate azomethine ylide and the
transition states of the 1,3-dipolar cycloaddition step showed that
the neucleophile and electrophile were simultaneously activated
by the bifunctional chiral bisphosphoric acids through formation
of hydrogen bonds. The effect of the bisphosphoric acids on the
reactivity and stereochemistry of three-component 1,3-dipolar
cycloaddition reaction was also theoretically rationalized. For the
1,3-dipolar cycloaddition reaction of azomethine ylide with
maleate, the two H-bonds of catalyst were opened to accept
maleate by its two hydroxyls forming two new H-bonds to
activate the electron-deficient CꢀC double bound. The ion pair
interaction between the phosphate of catalyst and the ester of
azomethine ylide accounted for the high stereoselectivity for this
1,3-dipolar cycloaddition reaction. For the dipolarophile, acrylate
or fumarate derivatives, one H-bond in catalyst remains to adjust
the acidity and basicity of two phosphoric acids to activate the
dipole and dipolarophile more effectively. These procedures
provide new accesses to a wide spectrum of structurally diverse
pyrrolidines and their potentially pharmaceutical relevant deri-
vatives with high enantiomeric purity.
Royer, J.; Husson, H.-P. Tetrahedron Lett. 1991, 32, 7531. (c) Anslow,
A. S.; Harwood, L. M.; Lilley, I. A. Tetrahedron: Asymmetry 1995, 6, 2465.
(d) Garner, P.; Ho, W. B.; Shin, H. J. Am. Chem. Soc. 1993, 115, 10742.
(e) Garner, P.; Dogan, O. J. Org. Chem. 1994, 59, 4. (f) Takano, S.;
Moriya, M.; Ogasawara, K. Tetrahedron: Asymmetry 1992, 3, 681. (g)
Wee, A. G. H. J. Chem. Soc., Perkin. Trans. 1 1989, 1363. (h) Nꢀajera, C.;
Gracia, R. M. D.; Sansano, J. M. Tetrahedron: Asymmetry 2006, 17, 1985.
(i) Nꢀajera, C.; Gracia, R. M. D.; Sansano, J. M.; Cossio, F. P. Eur. J. Org.
Chem. 2007, 5038.
(5) (a) Allway, P.; Grigg, R. Tetrahedron Lett. 1991, 32, 5817. (b)
Grigg, R. Tetrahedron: Asymmetry 1995, 6, 2475.
(6) (a) Longmire, J. M.; Wang, B.; Zhang, X. J. Am. Chem. Soc. 2002,
124, 13400. (b) Gothelf, A. S.; Gothelf, K. V.; Hazell, R. G.; Jørgensen,
K. A. Angew. Chem., Int. Ed. 2002, 41, 4236.
(7) (a) Chen, C.; Li, X.; Schreiber, S. L. J. Am. Chem. Soc. 2003,
125, 10174. (b) Kn€opfel, T. F.; Aschwanden, P.; Ichikawa, T.; Watanabe,
T.; Carreira, E. M. Angew. Chem., Int. Ed. 2004, 43, 5971. (c) Cabrera, S.;
Arrayꢀas, R. G.; Carretero, J. C. J. Am. Chem. Soc. 2005, 127, 16394. (d)
Yan, X.-X.; Peng, Q.; Zhang, Y.; Zhang, K.; Hong, W.; Hou, X.-L.; Wu,
Y.-D. Angew. Chem., Int. Ed. 2006, 45, 1979. (e) Zeng, W.; Chen, G.-Y.;
Zhou, Y.-G.; Li, Y.-X. J. Am. Chem. Soc. 2007, 129, 750. (f) Saito, S.;
Tsubogo, T.; Kobayashi, S. J. Am. Chem. Soc. 2007, 129, 5364. (g)
Tsubogo, T.; Saito, S.; Seki, K.; Yamashita, Y.; Kobayashi, S. J. Am. Chem.
Soc. 2008, 130, 13321. (h) Wang, C.-J.; Liang, G.; Xue, Z.-Y.; Gao, F.
J. Am. Chem. Soc. 2008, 130, 17250. (i) Arai, T.; Mishiro, A.; Yokoyama,
N.; Suzuki, K.; Sato, H. J. Am. Chem. Soc. 2010, 132, 5338. (j) Padilla, S.;
Tejero, R.; Adrio, J.; Carretero, J. C. Org. Lett. 2010, 12, 5608. (k) Zhang,
C.; Yu, S.-B.; Hu, X.-P.; Wang, D.-Y.; Zheng, Z. Org. Lett. 2010, 12, 5542.
(l) Yamashita, Y.; Guo, X.-X.; Takashita, R.; Kobayashi, S. J. Am. Chem.
Soc. 2010, 132, 3262. (m) Oura, I.; Shimizu, K.; Ogata, K.; Fukuzawa, S.
Org. Lett. 2010, 12, 1752. (n) Rocío, R. M.; María, G. E.; Adrio, J.;
Carretero, J. C. J. Org. Chem. 2010, 75, 233. (o) Rocío, R. M.; Alonso, I.;
Adrio, J.; Carretero, J. C. Chem.—Eur. J. 2010, 16, 5286.
(8) Kudryavtsev, K. V.; Zagulyaeva, A. A. Russ. J. Org. Chem. 2008,
44, 378.
’ ASSOCIATED CONTENT
ꢀ
(9) (a) Vicario, J. L.; Reboredo, S.; Badιa, D.; Carrillo, L. Angew.
Chem., Int. Ed. 2007, 46, 5168. (b) Ibrahem, I.; Rios, R.; Vesely, J.;
Cordova, A. Tetrahedron. Lett. 2007, 48, 6252. For a leading literature on
LUMO lowering by Iminium: (c) Jen, W. S.; Wiener, J. J. M.; MacMillan,
D. W. C. J. Am. Chem. Soc. 2000, 122, 9874.
(10) (a) Xue, M.-X.; Zhang, X.-M.; Gong, L.-Z. Synlett 2008, 691.
(b) Liu, Y.-K.; Liu, H.; Du, W.; Yue, L.; Chen, Y.-C. Chem.—Eur. J. 2008,
14, 9873.
S
Supporting Information. Experimental details, charac-
b
terization of new compounds, selected NMR and HPLC spectra,
and complete ref 24. This material is available free of charge via
the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
(11) Jiao, P.; Nakashima, D.; Yamamoto, H. Angew. Chem., Int. Ed.
2008, 47, 2411.
(12) For the preliminary report, see:Chen, X.-H.; Zhang, W.-Q.;
Gong, L.-Z. J. Am. Chem. Soc. 2008, 130, 5652.
Corresponding Author
gonglz@ustc.edu.cn; luosw@ustc.edu.cn
(13) For reviews:(a) Akiyama, T. Chem. Rev. 2007, 107, 5744. (b)
Doyle, A. G.; Jacobsen, E. N. Chem. Rev. 2007, 107, 5713. (c) Terada, M.
Chem. Commun. 2008, 4097. (d) Terada, M. Synthesis 2010, 12, 1929.
For leading references: (e) Uraguchi, D.; Terada, M. J. Am. Chem. Soc.
2004, 126, 5356. (f) Akiyama, T.; Itoh, J.; Yokota, K.; Fuchibe, K. Angew.
Chem., Int. Ed. 2004, 43, 1566. For recent examples not included in
reviews: (g) Jain, P.; Antilla, J. C. J. Am. Chem. Soc. 2010, 132, 11884. (h)
Husmann, R.; Sugiono, E.; Mersmann, S.; Raabe, G.; Rueping, M.; Bolm, C.
Org. Lett. 2011, 13, 1044. (i) Rueping, M.; Brinkmann, C.; Antonchick,
’ ACKNOWLEDGMENT
We are grateful for financial support from NSFC (20732006,
21072181), MOST (973 program 2009CB825300), Ministry of
Education, and CAS. We also thank Dr. Fei Liu in Australia for
polishing English and professor Terada (Tohoku University in
Japan) for donating extra pure silica gel (Merck).
’ REFERENCES
(1) Schreiber, S. L. Science 2000, 287, 1964.
ꢁ
A. P.; Atodiresei, I. Org. Lett. 2010, 12, 4604. (j) Coriꢀc, I.; M€uller, S.; List, B.
J. Am. Chem. Soc. 2010, 132, 17370. (k) Lu, G.-J.; Birman, V. B. Org. Lett.
(2) For reviews: Pearson, W. H. In Studies in Natural Product
Chemistry; Atta-Ur-Rahman, Ed.; Elsevier: New York, 1998; Vol. 1, p
323.
2011, 13, 356.
(14) Yu, J.; He, L.; Chen, X.-H.; Song, J.; Chen, W.-J.; Gong, L.-Z.
Org. Lett. 2009, 11, 4946.
(3) For reviews:(a) Sardina, F. J.; Rapoport, H. Chem. Rev. 1996,
96, 1825. (b) Coldham, I.; Hufton, R. Chem. Rev. 2005, 105, 2765. (c)
Pandey, G.; Banerjee, P.; Gadre, S. R. Chem. Rev. 2006, 106, 4484. (d)
Nair, V.; Suja, T. D. Tetrahedron 2007, 63, 12247. (e) For highlights, see:
Nꢀajera, C.; Sansano, J. M. Angew. Chem., Int. Ed. 2005, 44, 6272.
(4) (a) Padwa, A.; Chen, Y.-Y.; Chlacchio, U.; Dent, W. Tetrahedron
1985, 41, 3529. (b) Deprez, P.; Rouden, J.; Chiaroni, A.; Riche, C.;
(15) Chen, X.-H.; Wei, Q.; Luo, S.-W.; Xiao, H.; Gong, L.-Z. J. Am.
Chem. Soc. 2009, 131, 13819.
(16) (a) Matsunaga, S.; Das, J.; Roels, J.; Vogl, E. M.; Yamamoto, N.;
Iida, T.; Yamaguchi, K.; Shibasaki, M. J. Am. Chem. Soc. 2000, 122, 2252.
(b) Majima, K.; Takita, R.; Okada, R.; Ohshima, T.; Shibasaki, M. J. Am.
Chem. Soc. 2003, 125, 15837. (c) Kumagai, N.; Matsunnga, S.; Kinoshita,
T.; Harada, S.; Okada, S.; Sakamoto, S.; Yamaguchi, K.; Shibasaki, M.
13517
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518

Journal of the American Chemical Society
ARTICLE
J. Am. Chem. Soc. 2003, 125, 2169. (d) Vogl, E. M.; Matsunaga, S.; Kanai,
M.; Iida, T.; Shibasaki, M. Tetrahedron Lett. 1998, 39, 7917.
(17) (a) Kurihara, K.; Yamamoto, Y.; Miyaura, N. Adv. Synth. Catal.
2009, 351, 260. (b) Robinson, J. E.; Taylor, R. J. K. Chem. Commun.
2007, 1617.
(18) We also evaluated the reaction catalyzed by calcium salt of 1a,
but no reaction occurred. Moreover, ICP-OES analysis indicated trace
amount of metals (see Supporting Information). The results sufficiently
rule out the effect of metal contamination on the stereoselectivity, which
has been found in other reactions: (a) Hatano, M.; Moriyama, K.; Maki,
T.; Ishihara, K. Angew. Chem., Int. Ed. 2010, 49, 3823. (b) Klussmann,
M.; Ratjen, L.; Hoffmann, S.; Wakchaure, V.; Goddard, R.; List, B. Synlett
2010, 2189. (c) Terada, M.; Kanomata, K. Synlett 2011, 1255.
(19) (a) Gao, W. Z.; Zhang, X. M.; Raghunath, M. Org. Lett. 2005,
7, 4241. (b) Alemparte, C.; Blay, G.; Jørgensen, K. A. Org. Lett. 2005,
7, 4569. (c) Nꢀajera, C.; Retamosa, M. G.; Sansano, J. M. Angew. Chem.,
Int. Ed. 2008, 47, 6055. (d) Kim, H. Y.; Shih, H. J.; Knabe, W. E.; Oh, K.
Angew.Chem., Int. Ed. 2009, 48, 7420.
(20) For more examples, see:Cheng, M.-N.; Wang, H.; Gong, L.-Z.
Org. Lett. 2011, 13, 2418.
(21) (a) Akiyama, T.; Morita, H.; Itoh, J.; Fuchibe, K. Org. Lett.
2005, 7, 2583. (b) Yamanaka, M.; Itoh, J.; Fuchibe, K.; Akiyama, T.
J. Am. Chem. Soc. 2007, 129, 6756. (c) Akiyama, T.; Morita, H.; Bachu,
P.; Mori, K.; Yamanaka, M.; Hirata, T. Tetrahedron. 2009, 65, 4950. (d)
Gridnev, I. D.; Kouchi, M.; Sorimachi, K.; Terada, M. Tetrahedron Lett.
2007, 48, 497. (e) Simon, L.; Goodman, J. M. J. Am. Chem. Soc. 2008,
130, 8741. (f) Simon, L.; Goodman, J. M. J. Am. Chem. Soc. 2009,
131, 4070. (g) Simon, L.; Goodman, J. M. J. Org. Chem. 2010, 75, 589.
(h) Marcelli, T.; Hammar, P.; Himo, F. Chem.—Eur. J. 2008, 14, 8562.
(i) Marcelli, T.; Hammar, P.; Himo, F. Adv. Synth. Catal. 2009, 351, 525.
(j) Yamanaka, M.; Hirata, T. J. Org. Chem. 2009, 74, 3266. (k) Shi, F.-Q.;
Song, B.-A. Org. Biomol. Chem. 2009, 7, 1292.
(22) Backe, A. D. J. Chem. Phys. 1983, 98, 5648.
(23) Gill, P. M. W.; Johnson, B. G.; Pople, J. A.; Frisch, M. J. Chem.
Phys. Lett. 1992, 197, 499.
(24) Frisch, M. J., et al. Gaussian 03, Revision D.03; Gaussian, Inc.:
Wallingford, CT, 2004.
(25) Bahmanyar, S.; Houk, K. N. J. Am. Chem. Soc. 2001, 123, 11273.
(26) Rao, Q.-Q.; Luo, S.-W.; Gong, L.-Z. Chin. Sci. Bull. 2010,
55, 1742.
13518
dx.doi.org/10.1021/ja204218h |J. Am. Chem. Soc. 2011, 133, 13504–13518