A theoretical and experimental study of the effects of silyl substituents in enantioselective reactions catalyzed by diphenylprolinol silyl ether

Article abstract of DOI:10.1002/chem.201403514

The effect of silyl substituents in diphenylprolinol silyl ether catalysts was investigated. Mechanistically, reactions catalyzed by diphenylprolinol silyl ether can be categorized into three types: two that involve an iminium ion intermediate, such as fo

Full text of DOI:10.1002/chem.201403514

DOI: 10.1002/chem.201403514
Full Paper
&
Organocatalysis
A Theoretical and Experimental Study of the Effects of Silyl
Substituents in Enantioselective Reactions Catalyzed by
Diphenylprolinol Silyl Ether
Yujiro Hayashi,*[a, b] Daichi Okamura,[a, b] Tatsuya Yamazaki,[b] Yasuto Ameda,[b]
Hiroaki Gotoh,[b] Seiji Tsuzuki,[c] Tadafumi Uchimaru,*[c] and Dieter Seebach[d]
Abstract: The effect of silyl substituents in diphenylprolinol
silyl ether catalysts was investigated. Mechanistically, reac-
tions catalyzed by diphenylprolinol silyl ether can be catego-
rized into three types: two that involve an iminium ion inter-
mediate, such as for the Michael-type reaction (type A) and
the cycloaddition reaction (type B), and one that proceeds
via an enamine intermediate (type C). In the Michael-type re-
action via iminium ions (type A), excellent enantioselectivity
is realized when the catalyst with a bulky silyl moiety is em-
ployed, in which efficient shielding of a diastereotopic face
of the iminium ion is directed by the bulky silyl moiety. In
the cycloaddition reaction of iminium ions (type B) and reac-
tions via enamines (type C), excellent enantioselectivity is
obtained even when the silyl group is less bulky and, in this
case, too much bulk reduces the reaction rate. In other
cases, the yield increases when diphenylprolinol silyl ethers
with bulky substituents are employed, presumably by sup-
pressing side reactions between the nucleophilic catalyst
and the reagent. The conformational behaviors of the imini-
um and enamine species have been determined by theoreti-
cal calculations. These data explain the effect of the bulki-
ness of the silyl substituent on the enantioselectivity and re-
activity of the catalysts.
Introduction
The Michael addition of aldehydes to nitroalkenes is a syn-
thetically useful reaction, in which near perfect enantioselectiv-
ity can be achieved by using the diphenylprolinol trimethylsilyl
(TMS) ether (1) as a catalyst.[3] We used this asymmetric Mi-
chael reaction as a key step in the recent one-pot synthesis of
(À)-oseltamivir,[6] in a one-pot synthesis of ABT-341,[7] and in
a three-pot synthesis of prostaglandin E1 methyl ester,[8] in
which excellent enantioselectivities were realized. On the other
hand, in reactions such as the formal aza [3+3] cycloaddition
reaction of a,b-unsaturated aldehydes and enecarbamates,[9]
The field of organocatalysis is developing rapidly, and many
useful organocatalysts with unique properties have been de-
vised and applied to a plethora of asymmetric catalytic reac-
tions.[1] Diarylprolinol silyl ether[2] is one of the privileged orga-
nocatalysts, developed concurrently by our group[3] and Jør-
gensen’s group[4] (Figure 1). It is an effective catalyst for reac-
tions involving both enamine[1] and iminium reactive inter-
mediates.[5]
[a] Prof. Dr. Y. Hayashi, D. Okamura
Department of Chemistry, Graduate School of Science
Tohoku University, Aoba-ku, Sendai 980-8578 (Japan)
E-mail: yhayashi@m.tohoku.ac.jp
[b] Prof. Dr. Y. Hayashi, D. Okamura, T. Yamazaki, Y. Ameda, H. Gotoh
Previous address before June 30, 2012:
Department of Industrial Chemistry
Faculty of Engineering, Tokyo University of Science
Kagurazaka, Shinjuku-ku, Tokyo 162-8601 (Japan)
[c] Dr. S. Tsuzuki, Dr. T. Uchimaru
Nanosystem Research Institute
National Institute of Advanced Industrial Science and Technology
Tsukuba, Ibaraki 305-8568 (Japan)
E-mail: t-uchimaru@aist.go.jp
[d] Prof. Dr. D. Seebach
Laboratorium fꢀr Organische Chemie
Departement fꢀr Chemie und Angewandte Biowissenschaften, ETH-Zꢀrich
Hçnggerberg, Vladimir-Prelog-Weg 3, 8093 Zꢀrich (Switzerland)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201403514.
Figure 1. Organocatalysts examined in the present study.
Chem. Eur. J. 2014, 20, 1 – 13
1
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
the enantioselectivity is not perfect in the presence of diphe-
nylprolinol TMS ether, and diphenylprolinol tert-butyldimethyl-
silyl (TBS) ether was found to be a superior enantioselective or-
ganocatalyst relative to the trimethylsilylated analogue. In
these cases, the more bulky silyl group increased the enantio-
selectivity of the reaction.
bond, and the sc-exo conformation around the exocyclic CÀC
bond (Figure 2). The down conformation of the pyrrolidine ring
is also found for the crystal structure of this enamine. If the sc-
exo conformer also predominates in solution, the bulkiness of
the silyl substituent is likely to affect the enantioselectivity
greatly, because of its orientation and proximity to the reacting
trigonal center.
Recently, Seebach and co-workers determined the X-ray crys-
tal structures of iminium salts generated from cinnamaldehyde
and the diphenylmethyl- and trimethylsilylated diphenylproli-
nol ethers 2 and 1, respectively (Figure 1).[10] In the structures
of the iminium ions, there could be three conformations
around the exocyclic bond, the C=N and the C=C bond could
have an E or Z configuration, and the single bond between
C=C and C=N could have an s-cis or s-trans conformation: in
addition, there is puckering of the pyrrolidine ring. The pyrroli-
dine ring may adopt two distinct conformations, the down-
and up-puckered conformations. These conformations are de-
fined as conformations in which the Cg atom and the large
substituent at the Ca atom are located on the same and oppo-
site sides, respectively, of the plane defined by Cd, N, and Ca
atoms[11] (Figure 2). For both iminium ions derived from 1 and
2, X-ray crystallographic analysis shows an E/E configuration,
and the conformation around the exocyclic CÀC bond in the
Although the X-ray crystallographic analyses provided the
solid-state structures of the iminium ions 3, 4 and of the enam-
ine 5, and taking into account some preliminary calculations
and previously published NMR analyses,[10] the detailed overall
conformational space and dynamic behaviors of these species
still remain unclear. There are a couple of reports with diphe-
nylmethyl silyl ether 2 as a catalyst: Zhang and Liu[12] and
Pihko et al.[13] reported moderate enantioselectivity in the Mi-
chael reaction of malonate and isatylidene-3-acetaldehyde, and
in the Mukaiyama–Michael reaction of methacrolein and 5-
methylsiloxyfuran, respectively, whereas Melchiorre and co-
workers[14] obtained excellent enantioselectivity in the g-alkyla-
tion of a branched enal. There have been no systematic inves-
tigations into the substituent effects of the silyl group upon
conformational preference. Herein, we describe in detail our
computational and experimental investigations into these con-
formational and steric effects.
Results
Computational analysis
We carried out conformational analyses of the two iminium
ions 3 and 4 derived from diphenylprolinol TMS ether (1) and
cinnamaldehyde, and diphenylprolinol diphenylmethylsilyl
ether (2) and cinnamaldehyde, respectively (Figure 2). In addi-
tion, the conformations of the enamine 5 derived from diphe-
nylprolinol TMS ether (1) and phenylacetaldehyde were investi-
gated (Figure 2). Solid-state structures of the two iminium ions,
and of the enamine, have been obtained by X-ray crystal struc-
ture analysis.[10] Two structures with very small differences were
found in the unit cell of the crystal lattice for the iminium salts
3 (Figure 3a and b) and 4 (Figure 3e and f), and puckered con-
formers were seen in the crystal structure of the enamine 5
(Figure 3i). Starting from the X-ray crystal structures, we per-
formed conformational searches using CONFLEX 7[15] with the
MMFF94s force field[16] and a search limit of 20.0 kcalmolÀ1. For
the structures with relative energies with respect to the
lowest-energy structure that were calculated to be less than
3.0 kcalmolÀ1, we carried out DFT geometry optimizations at
the B3LYP/6-31G(d) level.[17] A factor of 0.9806 was used to cor-
rect the B3LYP/6-31G(d)-calculated zero-point energies.[18] Fur-
thermore, the energies of the optimized structures were evalu-
ated with single-point calculations at the M06-2X/6-311+
G(2df,2p) level.[19] Based on the M06-2X/6-311+G(2df,2p) ener-
gies, we estimated the population of each conformer at 298 K.
For the isomers of iminium ions 3 and enamine 5, entropic
and solvation contributions were also considered. The calculat-
ed values for entropic and solvation contributions are given in
the Supporting Information. After taking the entropic and sol-
Figure 2. Top: Schematic representations of synclinal-exo (sc-exo), synclinal-
endo (sc-endo), and antiperiplanar (ap) conformations of the exocyclic CÀC
bond of the iminium ion and the enamine. Middle: The up- and down-puck-
ered conformations of the pyrrolidine ring. Bottom-left: Schematic represen-
tation of the solid-state structure of the iminium ions of diarylprolinol silyl
ether. Bottom-right: Schematic representation of the solid-state structure of
the enamines of diarylprolinol silyl ether. The stereochemical nomenclature
defined here is used throughout the paper.
solid state is found to be (+)-synclinal (sc-exo; Figure 2). Addi-
tionally, the single bond between the C=C and C=N double
bonds adopts an s-trans conformation in the solid state. For
the pyrrolidine ring, these iminium ions adopt the down-puck-
ered conformation in the solid state. Seebach and co-workers
also determined the solid-state structure of the enamine de-
rived from diphenylprolinol TMS ether 1 and phenylacetalde-
hyde.[10a] This enamine adopts an s-trans conformation about
the CÀN single bond, an E configuration of the C=C double
&
&
Chem. Eur. J. 2014, 20, 1 – 13
www.chemeurj.org
2
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Figure 3. Top: Structures of iminium ion 3: a,b) crystal structures; c,d) B3LYP-optimized structures (entries 1 and 2 in Table 1). Middle: Structures of iminium
ion 4: e,f) crystal structures; g,h) B3LYP-optimized structures (entries 1 and 2 in Table 2). Bottom: Structures of enamine 5: i) crystal structure containing puck-
ered isomers; j–l) B3LYP-optimized structures (entries 1, 2, and 4 in Table 3).
vation contributions into account, however, the relative ener-
gies of the isomers of iminium ions 3, as well as those of en-
amine 5, did not alter significantly compared with those ob-
tained from the gas-phase electronic energies, and the error
bar on the Gibbs free energy predictions is expected to be sig-
nificantly larger than that of the electronic energies.[13] Accord-
ingly, the relative energies (including the zero-point energies)
reported herein are those obtained from gas-phase electronic
energies at the B3LYP/6-31G(d) and M06-2X/6-311+G(2df,2p)
levels. The DFT calculations were carried out by using the
Gaussian 09 program package.[20]
between C=C and C=N double bonds, respectively. Some of
the twelve conformers converged to the same structure during
B3LYP geometry optimizations, and five of the B3LYP-opti-
mized conformers remained unique. The two most stable con-
formers have (EC,N, EC,C) configurations for the C=N and C=C
double bonds. These are puckered isomers of the pyrrolidine
ring conformation: the most stable one adopts a down confor-
mation, whereas the second most stable one has an up confor-
mation (Figure 3c and d; Figures S1, S2, and S3a and b in the
Supporting Information). The population of these two con-
formers adopting the (EC,N, EC,C) configuration is estimated to
be 98%. The (ZC,N, EC,C) isomer was found to be the third most
stable configuration (Figures S1, S2, and S3c in the Supporting
Information). The B3LYP and M06-2X calculations suggest that
the (ZC,N, EC,C) configurational isomer is higher in energy by 3.6
and 2.2 kcalmolÀ1 than the most stable (EC,N, EC,C) configura-
tional isomer, and the population of the (ZC,N, EC,C) configura-
In Table 1 the results of the conformational analysis of the
iminium ion 3 are summarized. A conformational search of the
lowest-energy structure using the MMFF94s force field sug-
gested twelve conformers in the 3.0 kcalmolÀ1 energy range.
All these conformers have an sc-exo and an s-trans conforma-
tion around the exocyclic CÀC bond and the CÀC single bond
Chem. Eur. J. 2014, 20, 1 – 13
www.chemeurj.org
3
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Table 1. Results of conformational and configurational analysis of iminium ion 3 using the MMFF94s force field and subsequent DFT calculations.
Structure
CONFLEX
DFT calculation
M06-2X/6-311+G(2df,2p)
B3LYP/6-31G(d)
Relative
Steric
Relative
energy
Relative
energy
energy
energy
[kcalmolÀ1
Population
[%]
[kcalmolÀ1
]
[a.u.][a]
]
[a.u.][a]
[kcalmolÀ1
]
1
2
3
4
5
down
up
up
up
down
(EC,N, EC,C
(EC,N, EC,C
(ZC,N, EC,C
(EC,N, ZC,C
(EC,N, ZC,C
)
)
)
)
)
sc-exo
sc-exo
sc-exo
sc-exo
sc-exo
121.54
121.10
123.60
123.10
123.74
0.44
0.00
2.50
2.00
2.64
À1544.341302
À1544.341025
À1544.335649
À1544.332982
À1544.332837
0.00
0.17
3.55
5.22
5.31
À1544.190411
À1544.189181
À1544.186914
À1544.184190
À1544.184104
0.00
0.77
2.19
3.90
3.96
77
21
2
0.1
0.1
[a] A scaling factor of 0.9806 was used for the B3LYP/6-31G(d) zero-point energies. The zero-point energies are included.
Table 2. Results of conformational and configurational analysis of iminium ion 4 using the MMFF94s force field and subsequent DFT calculations.
Structure
CONFLEX
DFT calculation
M06-2X/6-311+G(2df,2p)
B3LYP/6-31G(d)
Relative
Steric
Relative
energy
Relative
energy
energy
energy
[kcalmolÀ1
Population
[%]
[kcalmolÀ1
]
[a.u.][a]
]
[a.u.][a]
[kcalmolÀ1
]
1
2
3
4
up
down
up
(EC,N, EC,C
(EC,N, EC,C
(EC,N, EC,C
)
)
)
sc-exo
sc-exo
sc-exo
sc-exo
162.71
163.05
165.61
165.27
0.00
0.33
2.90
2.56
À1927.704012
À1927.703936
À1927.703689
À1927.692497
0.00
0.05
0.20
7.23
À1927.523616
À1927.524789
À1927.521263
À1927.515647
0.00
À0.74
1.48
22
76
2
up
(EC,N, ZC,C
)
5.00
–
[a] A scaling factor of 0.9806 was used for the B3LYP/6-31G(d) zero-point energies. The zero-point energies are included.
tional isomer is calculated to be 2%. The calculations indicated
that the (EC,N, ZC,C) and (ZC,N, ZC,C) configurational isomers are
highly energetically disfavored, and contribution to the popu-
lation by these isomers must be negligible.
of the C=C double bond (Figures S1, S2, and S3g–i in the Sup-
porting Information). Isomers of the s-trans-Z, s-cis-E, and s-cis-
Z form were found to be energetically disfavored. The twelve
structures converged into nine structures during B3LYP geom-
etry optimizations. Regarding the conformation of the exocy-
clic CÀC bond, the enamine 5 has an sc-exo conformation in
the crystal structure, but B3LYP and M06-2X calculations, as
well as the force field calculations, suggest that the most
stable conformer would possess the ap conformation. The
populations of the ap and sc-exo conformers were calculated
to be almost equal (48 versus 51%). The sc-endo conformer
was also found in the energy range of 3.0 kcalmolÀ1 by the
MMFF94s conformational search, but B3LYP and M06-2X calcu-
lations suggest that the energy of the sc-endo structure is 3.2–
3.4 kcalmolÀ1 higher than that of the lowest-energy ap confor-
mer. For the structure of the pyrrolidine ring, DFT calculations,
as well as MMFF94s predictions, indicate that the down confor-
mation tends to be energetically more preferable than the up
conformation.
Table 2 shows the results of the conformational analysis of
the iminium ion 4. Four conformers with the sc-exo and s-trans
conformations were found in an energy range of 3.0 kcalmolÀ1
of the global minimum of the MMFF94s force field. Two iso-
mers, the energies of which were comparable, were found to
be the most stable conformers for the iminium ion 4. In
a slightly higher energy region, one more conformer was
found. These three conformers all adopt (EC,N, EC,C) configura-
tions and sc-exo conformations, and their structures are very
close to one another (Figures S1, S2, and S3d–f in the Support-
ing Information). M06-2X calculations suggest that the isomer
adopting a down conformation for the pyrrolidine ring is lower
in energy than the up conformers. The (EC,N, ZC,C) configuration-
al isomer was calculated to be 7.2 and 5.7 kcalmolÀ1 higher in
energy at the B3LYP and M06-2X levels, respectively, relative to
the most stable (EC,N, EC,C) isomer. The (ZC,N) configurational iso-
mers were found to be still higher in energy. Consequently, the
population of (EC,N, ZC,C), (ZC,N, EC,C), and (ZC,N, ZC,C) configuration-
al isomers was found to be negligible, and the iminium ion 4
will most likely adopt (EC,N, EC,C) configurations and s-trans/sc-
exo conformations, exclusively.
The lowest-energy conformers of the iminium ions 3 and 4
have (EC,N, EC,C) configurations and s-trans/sc-exo conformations
(Figure 3c and g). The s-trans-(E)-sc-exo structure is also ob-
served in the second minimum arrangement of the enamine 5
(Figure 3k). The structures of these conformers are very close
to the solid-state structures (Figure 3a, b, e, f, i). However, the
lowest-energy conformer of enamine 5 was found to adopt an
ap form. These conformational results indicate that the strong
conformational sc-exo preferences observed in iminium ions 3
and 4 are weakened in the enamine intermediate 5,[21] that is,
the gauche effect is stronger with the positively charged imini-
Table 3 gives the results of the conformational analysis of
the enamine 5. MMFF94s conformational search provided
twelve conformers in the energy range 3.0 kcalmolÀ1 from the
lowest-energy conformer. All the conformers have an s-trans
conformation of the CÀN single bond and an E configuration
&
&
Chem. Eur. J. 2014, 20, 1 – 13
www.chemeurj.org
4
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Table 3. Results of conformational and configurational analysis of enamine 5 using the MMFF94s force field and subsequent DFT calculations.
Conformation
CONFLEX
DFT calculation
B3LYP/6-31G(d)
[a.u.][a]
M06-2X/6-311+G(2df,2p)
Relative
Steric
energy
Relative
energy
]
Relative
energy
[kcalmolÀ1
energy
Population
[%]
[kcalmolÀ1
]
[a.u.][a]
[kcalmolÀ1
]
1
2
3
4
5
6
7
8
9
down
down
up
down
up
up
up
down
up
s-trans-(E)
s-trans-(E)
s-trans-(E)
s-trans-(E)
s-trans-(E)
s-trans-(E)
s-trans-(E)
s-trans-(E)
s-trans-(E)
ap
sc-exo
ap
sc-exo
sc-exo
sc-exo
sc-exo
sc-exo
sc-endo
152.65
153.91
154.40
152.82
153.51
153.81
153.49
153.22
154.60
0.00
1.26
1.75
0.17
0.86
1.16
0.84
0.57
1.95
À1505.844433
À1505.844032
À1505.843394
À1505.843267
À1505.843075
À1505.842023
À1505.841824
À1505.841145
À1505.839265
0.00
0.25
0.65
0.73
0.85
1.51
1.64
2.06
3.24
À1505.715861
À1505.715282
À1505.712704
À1505.714976
À1505.713217
À1505.710704
À1505.713334
À1505.713056
À1505.710523
0.00
0.36
1.98
0.56
1.66
3.24
1.59
1.76
3.35
46
25
2
18
3
0.2
3
2
0.2
[a] A scaling factor of 0.9806 was used for the B3LYP/6-31G(d) zero-point energies. The zero-point energies are included.
um nitrogen atom than with the neutral enamine nitrogen
atom.[22] It is also apparent from these calculations that there
are no electronic interactions (for example, silyl hypervalent in-
teractions) between the silyl atom and the p orbital of the
double bond in the iminium ions 3 and 4, or in the enamine 5.
The effect of the silyl substituent on the enantioselectivity
should be steric. However, the distances between the oxygen
atom and iminium moiety in iminium ions 3 and 4 are compa-
rable to or somewhat shorter (2.83–2.95 ꢁ) than the sum of
the van der Waals radii. Similarly, the distances between the
oxygen atom and enamine nitrogen atom are rather short
(2.85–2.98 ꢁ) in enamine 5. Thus, there are possible additional
electronic interactions between the oxygen atom and the imi-
nium moiety/enamine nitrogen atom.
ed the product in 90.0:10.0 e.r.[23a] Rios et al. used 20 mol% di-
phenylprolinol TMS ether (1) in toluene at 48C, which provided
good yields and excellent enantioselectivities.[23b]
Before investigating changes in the silyl substituents of the
catalyst, we first optimized the conditions for the Michael addi-
tion of bis(phenylsulfonyl)methane to crotonaldehyde, by
using the TMS catalyst 1 [Table 4, Eq. (1)]. The reaction was
performed in several solvents at room temperature for 20 h in
the presence of catalyst 1. The best yield and enantioselectivity
were observed in toluene. The absolute configuration (R) of
the product was assigned previously.[23]
Having found suitable reaction conditions, the effect of the
substituents on the silyl group of the diphenylprolinol organo-
catalyst was investigated [Table 5, Eq. (2)]. The enantioselectivi-
Keeping the structural information in the solid phase and
gas phase in mind, the effects of the substituents on the silyl
group in overall enantioselective catalytic reactions were inves-
tigated next.
Table 4. Solvent effect in the reaction of crotonaldehyde and bis(phenyl-
sulfonyl)methane catalyzed by 1.[a]
Michael addition of bis(phenylsulfonyl)methane to a,b-unsa-
turated aldehydes
Preliminary synthetic studies gave moderate enantioselectivity
using the diphenylprolinol TMS ether (1); therefore, we elected
to study the Michael addition of bis(phenylsulfonyl)methane to
trans-a,b-unsaturated aldehydes. Moreover, this is a synthetical-
ly useful reaction because bis(phenylsulfonyl)methane is a syn-
thetic equivalent of the methyl group. At the start of this proj-
ect, there were no reports of this reaction type being catalyzed
by organocatalysts; however, three independent groups subse-
quently reported similar reactions using the diphenylprolinol
silyl ether 1.[23] Palomo et al. used 1,3-benzodithiole tetraoxide
as a nucleophile and diphenylprolinol TMS ether (1) as a cata-
lyst. Although excellent enantioselectivity was obtained with
cinnamaldehydes, enantioselectivity decreased with aliphatic
enals (90.0:10.0 enantiomeric ratio (e.r.) in the reaction of cro-
tonaldehyde).[23c] Aleman and co-workers reported the reaction
using diphenylprolinol TMS ether (1) in the presence of LiOAc
in THF, and excellent enantioselectivities were obtained in
most of the reactions except for crotonaldehyde, which afford-
Solvent
Yield [%][b]
e.r.[c]
1
2
3
4
5
CH3CN
MeOH
Et2O
CH2Cl2
toluene
35
19
58
60
90
72.5:27.5
79.0:21.0
83.0:17.0
77.0:23.0
85.5:14.5
[a] Reaction conditions: crotonaldehyde (1.5 mmol), bis(phenylsulfonyl)-
methane (0.5 mmol), catalyst 1 (0.05 mmol), solvent (1.0 mL), room tem-
perature, reaction time 20 h. [b] Yield of purified product. [c] Enantiomeric
ratio (e.r.) was determined by HPLC analysis on a chiral phase after reduc-
tion to the corresponding alcohol with NaBH4.
Chem. Eur. J. 2014, 20, 1 – 13
www.chemeurj.org
5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Table 5. The effect of the silyl substituents of the catalyst in the reaction
of crotonaldehyde with bis(phenylsulfonyl)methane.[a]
R1
R2
Temp. [8C]
Yield [%][b]
e.r.[c]
1
2
3
4
5
6
7[d]
8
Me3Si
Et3Si
H (1)
H
H
H
H (2)
H
H (2)
H
CF3
CF3
23
23
23
23
23
23
0
0
23
23
90
83
92
78
88
90
94
87
15
12
85.5:14.5
86.5:13.5
89.5:10.5
90.5:9.5
91.5:8.5
92.0:8.0
95.0:5.0
94.5:5.5
90.0:10.0
92.5:7.5
tBuMe2Si
PhMe2Si
Ph2MeSi
Ph3Si
Ph2MeSi
Ph3Si
Figure 4. Relationship of yield and enantioselectivity versus time in the Mi-
chael reaction of crotonaldehyde with the bis-sulfone catalyzed by 1. The re-
action was performed by using crotonaldehyde (1.8 mmol), bis-sulfone
(0.6 mmol), catalyst 1 (0.06 mmol), and toluene (1.2 mL) at 08C. Dark gray
line: enantiomeric ratio of product. Light gray line: yield of product.
9
10[d]
Me3Si
Et3Si
[a] Reaction conditions: unless noted otherwise, the reaction was per-
formed using crotonaldehyde (1.5 mmol), bis-sulfone (0.5 mmol), organo-
catalyst (0.05 mmol), and toluene (1.0 mL) at room temperature for 20 h.
[b] Yield of purified product. [c] Enantiomeric ratio (e.r.) was determined
by chiral-phase HPLC analysis after reduction to the corresponding alco-
hol with NaBH4. [d] The reaction was performed at 08C for 48 h.
ty with the DPMS catalyst 2 was found to be significantly
higher than the corresponding reactions with the TMS ether
catalyst 1 (Table 6, entries 1–4). When the b-substituent of the
acrolein is an electron-withdrawing group, such as ethoxycar-
bonyl, a higher enantioselectivity was observed with the DPMS
ether catalyst 2 (Table 6, entry 5). From these results, it is con-
cluded that the conjugate addition of bis(phenylsulfonyl)me-
thane to a,b-unsaturated aldehydes under kinetic control pro-
ceeds with higher enantioselectivity when a prolinol catalyst
with bulkier silyl substituents is used. Experimentally, we found
ty obtained with the triethylsilyl (TES) catalyst was found to be
the same as that with the TMS version (86.5:13.5 e.r., Table 5,
entries 1 and 2). When the substituents on the silyl group
became more bulky, such as with the TBS group, the enantio-
selectivity increased somewhat (89.5:10.5 e.r., Table 5, entry 3).
In the case of the dimethylphenylsilyl ether, an e.r. of 90.5:9.5
was obtained. Good enantioselectivity (92.0:8.0 e.r.)
was also obtained with the diphenylmethylsilyl
(DPMS) ether (2; Table 5, entry 5) and the triphenylsil-
Table 6. Comparison of TMS ether catalyst 1 and DPMS ether catalyst 2 in the Michael
reaction of b-substituted a,b-unsaturated aldehydes with bis(phenylsulfonyl)meth-
ane.[a]
yl[24] ether (Table 5, entry 6). The enantioselectivity in-
creased to 95.0:5.0 e.r. when the reaction was carried
out at a lower temperature (08C, Table 5, entry 7).
Notably in the reaction catalyzed by bis-trifluoro-
methyl-substituted diarylprolinol silyl ether, the enan-
tioselectivity was high, but the yield dropped
(Table 5, entries 9 and 10). Thus, the best catalyst in
this reaction was found to be DPMS ether (2).
R
Catalyst 1
Time [h] Yield [%][b] e.r.[c]
Catalyst 2
Time [h] Yield [%][b] e.r.[c]
Next, the yield and enantioselectivity versus the re-
action time were investigated for this Michael addi-
tion catalyzed by diphenylprolinol TMS ether 1. As
shown in Figure 4, the yield was observed to gradual-
ly increase, whereas the enantioselectivity remained
constant throughout the reaction. These results indi-
cate that no retro Michael reaction occurred in the
course of the reaction and the enantioselectivity is
determined by kinetic control.
1[d] Me
2[e] nPr
3[e] iBu
20
8
17
94
82
77
85
65
90.0:10.0 48
90.0:10.0 24
90.0:10.0 17
87
87
82
82
81
95.0:5.0
94.5:5.5
93.5:6.5
95.5:4.5
97.5:2.5
4[e] CH2CH2Ph 12
93.0:7.0
95.0:5.0
28
12
5
CO2Et
10
[a] Unless noted otherwise, the reaction was performed using a,b-unsaturated alde-
hyde (1.5 mmol), bis(phenylsulfonyl)methane (0.5 mmol), organocatalyst (0.05 mmol),
and toluene (1.0 mL) at room temperature. [b] Yield of purified product. [c] Enantio-
meric ratio (e.r.) was determined by HPLC analysis on a chiral phase after reduction to
the corresponding alcohol by treatment with NaBH4. [d] The reaction was performed
at 08C. [e] Catalyst (0.1 mmol) was employed.
Comparison of the two catalysts, that is, TMS ether
(1) and DPMS ether (2), was further investigated in
the Michael addition of the bis(phenylsulfonyl)me-
thane with a selection of b-substituted, trans-a,b-un-
saturated aldehydes [Eq. (3)], and the results are col-
lected in Table 6. When the b-substituent of the acro-
lein is nPr, iBu, or 2-phenylethyl, the enantioselectivi-
&
&
Chem. Eur. J. 2014, 20, 1 – 13
www.chemeurj.org
6
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
the diphenylprolinol DPMS ether
(2) to be the best organocata-
lyst.
Table 7. Effect of silyl organocatalyst in the carbo [3+3] cyclization of cinnamaldehyde and dimethyl 3-oxopen-
tanedioate.[a]
In general, the reactivity of
the DPMS ether 2 is lower than
that of the TMS ether 1, and
longer reaction times are re-
quired for 2 (Table 6). However,
it should be emphasized that
the yield was higher when the
more bulky catalyst 2 was em-
ployed in the case of the more
reactive Michael acceptor (see
Table 6, entry 5). Such data pro-
vide evidence compatible with
the assumption that the catalyst
1 might add to the more reac-
tive Michael acceptor to form
the adduct 6, from which the
catalyst is released slowly by
elimination (see section 2.1.5.1 in
ref. [21]). This side reaction
might be less pronounced when
a more bulky silyl ether is em-
ployed.
Catalyst
Time [min]
Yield [%][b]
e.r.[c]
1
2
3
R1 =TMS, R2 =H (1)
R1 =TBS, R2 =H
R1 =Ph2MeSi, R2 =H (2)
50
60
80
79
75
76
95.5:4.5
97.5:2.5
97.5:2.5
[a] The reaction was performed using cinnamaldehyde (0.5 mmol), dimethyl 3-oxopentanedioate (0.56 mmol),
organocatalyst (0.05 mmol), benzoic acid (0.1 mmol), and CH2Cl2 (1.0 mL) at room temperature. [b] Yield of puri-
fied product. [c] Enantiomeric ratio (e.r.) was determined by HPLC analysis on a chiral phase after conversion to
the corresponding silyl ether by treatment with TMSCl and imidazole.
Table 8. Effect of silyl organocatalyst in the one-pot synthesis of a bicyclo[3.3.0]octane derivative.[a]
Comparison of the selectivity
of catalysts 1 and 2 in some
other enantioselective reac-
tions involving iminium ions as
reactive intermediates
Catalyst
Time [h][b]
Yield [%][c]
e.r.[d]
1
2
3[e]
R1 =TMS, R2 =H (1)
R1 =TBS, R2 =H
R1 =Ph2MeSi, R2 =H (2)
25
20
6.5
58
63
52
94.0:6.0
97.5:2.5
97.5:2.5
Having determined the superior-
ity of the DPMS ether 2 over the
TMS ether 1 in the asymmetric
Michael addition of bis(phenyl-
sulfonyl)methane to a,b-unsatu-
rated aldehydes, we investigated
the efficiency of 2 in other reac-
tions, in which the organocata-
lyst 1 had been employed and
studied previously. First, the
effect of the catalyst was investi-
gated in the carbo [3+3] cycliza-
[a] The reaction was performed using cinnamaldehyde (0.5 mmol), cyclopentadiene (1.5 mmol), organocatalyst
(0.05 mmol), p-nitrophenol (0.1 mmol), and MeOH (1.0 mL) at room temperature. After the first addition,
iBu2NH (0.5 mmol) and p-nitrophenol (0.5 mmol) were added and the reaction was further stirred for one day
at room temperature. [b] Reaction time for the first Michael reaction. [c] Yield of purified product. [d] Enantio-
meric ratio (e.r.) was determined by HPLC analysis on a chiral phase. [e] Catalyst 2 (0.1 mmol) was employed.
to 97.5:2.5 e.r., when bulky TBS ether and DPMS ether (2) cata-
tion of cinnamaldehyde and dimethyl 3-oxopentanedioate
through a domino Michael addition/Knoevenagel condensa-
tion, in which the key intermediate is an iminium ion [Table 7,
Eq. (4)].[25] When TMS ether 1 was employed as the catalyst, an
e.r. of 95.5:4.5 was obtained. A higher enantioselectivity
(97.5:2.5 e.r.) was indeed obtained with the more bulky TBS
catalyst and DPMS ether catalyst 2.
We then evaluated our previous one-pot synthesis of
a bicyclo[3.3.0]octatriene derivative involving Michael addition
and intramolecular addition of the cyclopentadiene moiety
and dehydration [Table 8, Eq. (5)].[26] Good enantioselectivity
(94.0:6.0 e.r.) was obtained when the TMS ether catalyst 1 was
employed. Again, the enantioselectivity increased, in this case
lysts were used.
Next, the Michael addition of nitromethane to cinnamalde-
hyde was compared [Table 9, Eq. (6)].[27] Excellent enantioselec-
tivity had already been obtained by the use of the TMS ether
catalyst 1 (97.5:2.5 e.r.). The enantioselectivity (98.5:1.5 e.r.) was
similar when the DPMS ether 2 was employed. Each reaction
was examined three times; the error was found to be within
1%. In this reaction, the bulkiness of the silyl substituent did
not affect the enantioselectivity significantly and both TMS
ether catalyst 1 and the DPMS ether catalyst 2 gave excellent
enantioselectivity.
We have now also investigated the Diels–Alder reaction of
cinnamaldehyde with cyclopentadiene [Table 10, Eq. (7)].[28]
Chem. Eur. J. 2014, 20, 1 – 13
www.chemeurj.org
7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Comparison of catalysts 1 and 2 in some reactions involving
enamines as reactive intermediates
Table 9. Effect of silyl organocatalyst in the Michael addition of nitrome-
thane to cinnamaldehyde.[a]
We also studied enantioselective reactions involving enamine
intermediates, rather than iminium ions derived from catalysts
1 and 2. A case in point is the Michael addition of propanal to
nitrostyrene [Table 11, Eq. (8)]. Excellent enantioselectivities
were reported with the TMS ether catalyst 1.[3,29] With the
Catalyst
Time [h]
Yield [%][b]
e.r.[c]
1
2
R1 =TMS, R2 =H (1)
R1 =Ph2MeSi, R2 =H (2)
16
17
90
90
97.5:2.5
98.5:1.5
[a] The reaction was performed using cinnamaldehyde (0.5 mmol), nitro-
methane (1.5 mmol), organocatalyst (0.05 mmol), benzoic acid (0.1 mmol),
and MeOH (1.0 mL) at room temperature. [b] Yield of purified product.
[c] Enantiomeric ratio (e.r.) was determined by HPLC analysis on a chiral
phase.
Table 11. Effect of silyl organocatalyst in the Michael addition of propanal
to nitrostyrene.[a]
Catalyst
Time [h] syn:anti Yield [%][b] e.r.[c]
1[d] R1 =TMS, R2 =H (1)
2
5
16:1
>20:1
85
89
99.5:0.5
99.5:0.5
R1 =Ph2MeSi, R2 =H (2) 24
[a] The reaction was performed using nitrostyrene (1.0 mmol), propanal
(10 mmol), and organocatalyst (0.1 mmol) in hexane (1.0 mL) at 08C.
[b] Yield of purified product. [c] Enantiomeric ratio (e.r.) was determined
by HPLC analysis on a chiral phase. [d] Data taken from ref. [3].
Previously, we had reported the trifluoromethyl-substituted di-
arylprolinol silyl ether as an effective enantioselective catalyst
in this [4+2] cycloaddition; however, the yield was poor in the
reaction catalyzed by diphenylprolinol silyl ether. The effect of
the silyl substituent on the enantioselectivity with catalysts
1 and 2 is shown in Table 10. Both catalysts gave similar en-
do:exo selectivities, and the same enantioselectivity for the
major exo isomer. This case possibly proceeds through a differ-
ent reaction mode from the other highly enantioselective reac-
tions (see above) that, likewise, involve iminium ions derived
from catalysts 1 and 2.
DPMS ether catalyst 2, the enantioselectivity was also excel-
lent, the syn selectivity increased, but a longer reaction time
was required.
We also studied the enamine-based Michael addition of an
aldehyde to N-phenylmaleimide [Table 12, Eq. (9)], a reaction
Table 10. Effect of silyl organocatalyst in the Diels–Alder reaction of cin-
namaldehyde with cyclopentadiene.[a]
Table 12. Effect of silyl catalyst in the Michael reaction of 3-methylbuta-
nal and N-phenylmaleimide.[a]
Catalyst
Time Yield[b] exo:endo[c]
e.r.[d]
endo
Catalyst
anti:syn
Yield [%][b]
e.r.[c]
[h]
[%]
exo
1[d]
2
R1 =TMS, R2 =H (1)
R1 =TMS, R2 =H (1)
R1 =Ph2MeSi, R2 =H (2)
8:1
7:1
5:1
70
65
38
99.5:0.5
99.5:0.5
99.5:0.5
1
2
3
4
R1 =TMS, R2 =H (1)
20
14
16
86
80
80:20
77:23
84:16
85:15
91.5:8.5 76.5:23.5
91.5:8.5 85.5:14.5
97.5:2.5 91.5:8.5
98.5:1.5 94.0:6.0
R1 =Ph2MeSi, R2 =H (2) 20
3
R1 =TMS, R2 =CF3
R1 =TES, R2 =CF3
20
20
[a] The reaction was performed using N-phenylmaleimide (0.5 mmol), 3-
methylbutanal (1.0 mmol), and organocatalyst (0.05 mmol) in CHCl3
(2.0 mL) at 238C for 24 h. [b] Yield of purified product. [c] Enantiomeric
ratio (e.r.) was determined by HPLC analysis on a chiral column after con-
version to a,b-unsaturated ester by treatment with Ph3P=CHCO2Et.
[d] Data extracted from ref. [30].
[a] The reaction was performed using cinnamaldehyde (1.0 mmol), cyclo-
pentadiene (3.0 mmol), organocatalyst (0.1 mmol), trifluoroacetic acid
(0.2 mmol), and toluene (2.0 mL) at room temperature. [b] Yield of puri-
fied product. [c] Diastereomer ratio was determined by 1H NMR spectros-
copy. [d] Enantiomeric ratio (e.r.) was determined by HPLC analysis on
a chiral phase.
&
&
Chem. Eur. J. 2014, 20, 1 – 13
www.chemeurj.org
8
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
that has been reported by Cordova et al. to give excellent
enantioselectivity with the TMS catalyst 1.[30] We carried out
the reaction using both catalyst 1 and 2, and found that both
gave excellent enantioselectivity, but the reactivity of catalyst
2 was lower than that of 1. Although Cordova and co-workers
proposed structure A as the transition-state model,[30] transi-
tion-state model B (shown with Table 12) was proposed by one
of the present authors (Y.H.) in the Michael reaction of a-alkox-
yaldehyde and N-(p-nitrophenyl)maleimide when the one-pot
synthesis of Tamiflu was investigated.[6d]
enantioselectivity is obtained with the sterically less
bulky silyl substituent, for example, with TMS ether 1.
Type C: Reactions involving an enamine intermediate. Smaller
silyl substituents on the catalyst, such as in TMS ether 1,
afford products with excellent enantioselectivities and
reasonable yields.
The effect of the silyl substituent on the enantioselectivity in
each category will be discussed by considering the structures
of the iminium ion and enamine, as revealed by our computa-
tional investigation. [Note: we refer to phenyl-substituted reac-
tive intermediates in the following discussions; other substitu-
ents, such as CH3, CH2-alkyl, CH(CH3)2, cyclohexyl, and CH=CH2,
on the iminium ion can lead to a reversal of the stereotopicity
descriptor according to CIP rules.]
As a last example of a reaction involving an enamine inter-
mediate, we highlight our previous work on attaining near-per-
fect enantioselectivity with the diphenylprolinol TMS ether 1 in
an intramolecular formal [6+2] cycloaddition [Eq. (10)].[31] Here,
there is no need to employ the bulky silyl ether 2.
In the reaction via iminium ions 3 and 4, which are derived
from cinnamaldehyde and the diphenylprolinol TMS ether (1)
or the diphenylprolinol DPMS ether (2) (Figure 2), respectively,
nucleophilic attack of the reagents occurs selectively from the
Si face of the electrophilic b-carbon of the iminium ion
(Figure 5). In the reaction via enamine 5, which is derived from
prolinol silyl ether 1 and phenylacetaldehyde, nucleophilic
attack occurs selectively from the Si face of the nucleophilic a-
carbon of the enamine 5 (Figure 5). These diastereo-differentia-
tions stem from two factors: 1) iminium ions 3 and 4 adopt an
(EC,N, EC,C) configuration and an s-trans conformation between
the C=C and C=N double bonds, and the enamine 5 preferen-
tially adopts an E configuration and s-trans conformation
around the exocyclic NÀC bond; and 2) the syn face of the pyr-
rolidine ring is blocked by bulky substituents ÀCPh2(OSiMe3)/À
CPh2(OSiPh2Me). Consequently, an Re-face approach of a nucle-
ophile to the iminium ions 3 and 4, and of an electrophile to
the enamine 5, will be sterically disfavored as compared to an
Si-face approach. These conclusions are well supported by our
conformational structure searches. The iminium ion 4 and en-
amine 5 were both calculated to exclusively adopt the s-trans-
(EC,N, EC,C) and (E, s-trans) structure, respectively (Tables 2 and
3). This conformational preference remains for the iminium ion
3, of which a small percentile (2% at 298 K, Table 1) was calcu-
lated to exist in the (ZC,N, EC,C) form (Table 1, see below). For the
exocyclic CÀC bond, our conformational analysis suggested ex-
clusive sc-exo conformations for the iminium ions 3 and 4
(Tables 1 and 2), as well as sc-exo or ap conformations for the
neutral enamine 5 (Table 3). In the sc-exo and ap conformers of
3, 4 and 5, one of the phenyl rings on the carbon atom is lo-
cated above the five-membered pyrrolidine ring (Figures S1
and S2 in the Supporting Information), and hence the top or
Discussion
Enantioselectivity
In general, the organocatalytic reaction of pyrrolidine-based
catalysts has previously been categorized by two types of reac-
tions: those involving neutral enamine intermediates and
those involving iminium ion intermediates. We propose those
reactions involving iminium ions to be subdivided into two
types: one involving Michael-type reaction modes and the
other involving cycloaddition reaction modes. There are thus
three types of reactions, and the effect of the silyl substituent
in each type will be discussed according to the following clas-
sifications:
Type A: Michael-type reactions involving an iminium ion in-
termediate. The catalyst with bulkier substituents gener-
ally gives better enantioselectivity; for instance, the di-
phenylprolinol DPMS ether 2 was found to be more se-
lective than the TMS ether 1. The exception is the Mi-
chael reaction of a,b-un-
saturated aldehyde with
nitromethane, in which
excellent enantioselectivi-
ty is obtained in both the
TMS ether catalyst 1 and
the DPMS catalyst 2.
Type B: Cycloaddition reactions
involving an iminium ion
intermediate.
Excellent Figure 5. Schematic approach of reactants for type A, type B, and type C processes.
Chem. Eur. J. 2014, 20, 1 – 13
www.chemeurj.org
9
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Re faces of these species are all sterically crowded, whereas
the bottom or Si faces are all relatively accessible (Figure S3 in
the Supporting Information).
tion to proceed through a transition state whereby the reac-
tive nucleophilic center (the Ca-position) is relatively close to
the pyrrolidine ring [Eq. (10)].[31] Accordingly, the less bulky silyl
substituent ÀCPh2(OSiMe3) can efficiently block attack from the
top p face of the reacting trigonal center of the enamine. No-
tably, contribution by an (E)-s-cis form was calculated to be
negligible for enamine 5. A less clear case or process of enan-
tioselection occurs in the Michael addition of aldehyde-derived
enamines to nitroalkenes. These reactions have been proposed
to proceed via putative zwitterionic, cyclobutane, and dihy-
drooxazine intermediates both on and off the catalytic cycle
(Table 11).[29] Although the rate-determining step is not clear,
the enantio-determining step is suggested to be the addition
of the enamine to the nitroalkene through the model shown
in Table 11. The chirality at both a- and b-positions of the
formyl group would arguably be determined in this addition
step, even though the a-position might undergo subsequent
isomerization.[29a] In this model, a small TMS ether is sufficient
to cover the Re face efficiently. In the Michael addition of alde-
hyde-derived enamines to N-phenylmaleimide, there is thus
the possibility of a [2+2] cycloaddition followed by ring open-
ing of a cyclobutane intermediate.[29f] In this case again, the
enantio-determining step would be an addition step [Table 12,
Eq. (9)]. For both such cycloaddition and Michael reactions,
very high enantioselectivity has been attained by incorporation
of a small silyl ether into the prolinol catalyst. In the Michael
reaction of aldehyde and nitroalkenes, even the much smaller
methyl ether of diphenylprolinol is known to promote the re-
action with excellent enantioselectivity.[32] Consequently, the
TMS ether catalyst 1 is expected to be sufficient in achieving
satisfactory diastereotopic p-facial differentiation for reactions
involving catalyst-derived enamines as intermediates (Figure 5;
type C).
However, the ability of catalyst 1 to achieve enantiomeric
differentiation is slightly lower than that of catalyst 2 in cases
of reactions involving iminium intermediates of type A
(Tables 5–8). Two factors are considered to marginally lower
the enantioselectivity for reactions mediated by catalyst 1.
First, a top-face attack on the electrophilic carbon atom (the
Cb-position) of the iminium ion 3 may not be fully blocked by
the silyl substituent ÀCPh2(OSiMe3). For the iminium ion 4, one
of the phenyl groups on the silicon atom is located on the
upper side, positioned near the electrophilic center of the imi-
nium portion. In addition to the phenyl ring on the carbon
atom, the phenyl ring on the silicon atom can block attack of
a reagent from the top face in the iminium ion 4. Compared
with iminium ion 4, the top face of iminium ion 3 is relatively
less hindered due to the lack of phenyl groups on the silicon
atom (Figure 3 and Figure S1 in the Supporting Information).
Second, contributions of a small population of the (ZC,N, EC,C
)
isomers of 3 could be responsible for lowering the overall
enantioselectivity, since a bottom facial attack of this diastereo-
meric iminium ion would result in the opposite (and minor)
enantiomeric product.
The Michael reaction of a,b-unsaturated aldehyde and nitro-
methane is an exception to type A trends, in which both the
TMS catalyst 1 and the DPMS catalyst 2 gave excellent enantio-
selectivity (Table 9). As there would be an ionic interaction be-
tween a cationic portion of iminium ion and an anionic moiety
of the nitronate ion, nucleophile and Michael acceptor should
approach each other as shown in Table 9. Thus, even small
TMS-substituted catalyst 1 affords an excellent enantioselectivi-
ty. It would be important to consider the trajectory of the re-
agent to evaluate the effectiveness of the silyl substituent of
the catalyst in the Michael reaction.
Reactivity
In reactions of type B, via iminium ion intermediates, both
the a- and b-positions of the a,b-unsaturated system partici-
pate. As silyl substituents, such as ÀCPh2(OSiMe3), chiefly shield
the top face of the a- and b-positions, excellent stereoselectiv-
ity is obtained with the TMS catalyst 1. Bulky substituents such
as ÀCPh2(OSiPh2Me) are not necessary for achieving high over-
all enantioselectivity for this type of catalytic process.
In general, reactions catalyzed by diphenylprolinol silyl ethers
with bulky substituents are slow. As shown in Table 6, the reac-
tion time increased when the DPMS ether catalyst 2 was em-
ployed, as compared with reactions catalyzed by the TMS
ether 1 (see Tables 7 and 8). The reaction outlined in Table 7
was completed within 50 min with the TMS ether catalyst 1,
whereas a longer reaction time (80 min) was necessary with
the DPMS ether catalyst 2. For the reaction shown in Table 8,
10 mol% of the TMS ether catalyst 1 was found to be sufficient
to promote the reaction, whereas a higher catalyst loading of
the bulkier DPMS derivative 2 (20 mol%) was required to
obtain the product in reasonable yield.
In reactions of type C, the p-selective facial addition of the
catalyst-derived chiral (Z)-enamine 5 to a nonchiral electrophile
is the initial diastereo-differentiating step, irrespective of the
next mechanistic steps being put forward. We reason that this
initial event, although not necessarily rate-determining, defines
the absolute and relative Ca (and Cb) configurations in all en-
suing intermediates. Currently, the exact details of interrelated
catalytic cycles, equilibria, and intermediates are still a matter
of debate.[29] It is nevertheless reasonable to suggest that this
initial addition step installs the necessary stereogenicity in sub-
sequent intermediates so as to influence the resultant Ca
enantioselectivity in the product after the chiral prolinol cata-
lyst is released. The intramolecular formal [6+2] cycloaddition
reaction of aldehyde and fulvene is a clear case to consider
first. Ab initio and DFT calculations thus revealed the cycloaddi-
In contrast to the general observations that shorter reaction
times result from reactions catalyzed by diphenylprolinol silyl
ether with smaller silyl substituents, we found a case in which
higher yields were observed when the bulkier catalyst 2 was
employed. As shown in entry 5 of Table 6, diphenylprolinol
DPMS ether catalyst 2 gave a higher yield of the Michael
adduct, which we attributed to the TMS ether 1 “overreacting”
with the ester-activated Michael acceptor, thereby forming the
unproductive species 6. In this particular case, the bulky sub-
&
&
Chem. Eur. J. 2014, 20, 1 – 13
www.chemeurj.org
10
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
stituent may suppress this side reaction and thereby increase
the yield. In the a-benzoyloxylation of an aldehyde,[33] the cata-
lyst with bulky silyl substituents also gave higher reactivity. In
this case, the gradual reaction of the catalyst with benzoyl per-
oxide was suppressed with the bulkier TBS ether catalyst
[Eq. (11)].
the double bond in the iminium ion or in the enamine, and
the predominant effects of the silyl substituents were steric.
Optimized structures of the most stable geometry of iminium
ion 4 indicate that the bulky silyl group CPh2(OSiPh2Me) shields
the Re face of the Cb-position efficiently, whereas less effective
shielding of the Cb-position results with the smaller silyl group
CPh2(OSiMe3) on the iminium ion 3. This is the main reason
why a higher enantioselectivity results for type A reactions
when a prolinol catalyst with a bulkier silyl moiety is employed.
A structural comparison of the iminium ions 3 and 4 leads to
the conclusion that the relative bulkiness of the substituent
CPh2(OSiPh2Me) will reduce the reactivity in some cases, but is,
at the same time, able to enhance the enantioselectivity of the
catalyst 2. In addition, the existence of an s-trans-(ZC,N, EC,C) dia-
stereoisomer of the iminium ion 3[10c] could possibly further
lower the enantioselectivity with the TMS silyl catalyst 1 in
type A reactions.
Conclusion
We have investigated the effects of the silyl substituents in di-
phenylprolinol silyl ether catalysts in several types of enantio-
selective reactions. We have defined three types of reactions
catalyzed by diphenylprolinol silyl ethers. Type A is a Michael-
type reaction of a,b-unsaturated aldehydes involving iminium
ion intermediates, in which a higher enantioselectivity is realiz-
ed when the catalyst with a bulkier silyl moiety is employed
except for the Michael reaction of an a,b-unsaturated aldehyde
with nitromethane. One of the best substituent patterns on
the silyl atom in the diphenylprolinol silyl ether catalyst is the
diphenylmethyl group. Type B is a cycloaddition reaction via
iminium ion intermediates, in which small substituents on the
silyl atom, such as the trimethylsilyl (TMS) group, affords excel-
lent enantioselectivity. Type C is a reaction involving enamine
intermediates derived from an aldehyde as a nucleophile. In
this reaction, small substituents on the silyl group provide ex-
cellent enantioselectivity. In general, a bulky silyl group in the
diphenylprolinol silyl ether catalyst retards the reaction for
steric reasons. In certain cases, however, a better yield results
because the bulky silyl substituents suppress unproductive
side reactions or destruction of the catalyst molecule that
become more prevalent with more nucleophilic diphenylproli-
nol silyl ether catalysts.
For type C reactions, the electrophilic reagent attacks the
Ca-position of the neutral enamine, whereas for type B reac-
tions, attack of the reagent occurs simultaneously at the Ca-
and Cb-positions of the iminium species (Figure 5). Optimized
structures of the most stable forms of the iminium ions 3 and
4, as well as that of the enamine 5, indicate that both substitu-
ents CPh2(OSiMe3) and CPh2(OSiPh2Me) can effectively shield
the Re face of the Ca-position. This finding is in agreement
with the fact that even relatively small substituents on the silyl
atom in the organocatalyst can infer excellent enantioselectivi-
ty for type B and C reactions.
The theoretical and experimental findings described herein
should provide valuable information not only for the design of
new organocatalytic reactions, but also for the optimization of
key asymmetric reactions using diphenylprolinol silyl ether cat-
alysts.
Considering the discussion of stereoelectronic effects[10c] and
Jørgensen’s recent results about di- and trienamine catalysis,[34]
the details of the stereoselectivity of these reactions may be
more complex than summarized herein.
Acknowledgements
Over all three reaction types, silyl substituent effects on the
resultant enantioselectivity are rationalized through conforma-
tional analysis of the iminium ions and of the enamine, and
the trajectory of the reagents. Theoretical conformational anal-
yses were carried out for the iminium ions 3 and 4, derived
from diphenylprolinol TMS ether (1) and cinnamaldehyde, or
from diphenylprolinol DPMS ether (2) and cinnamaldehyde, as
well as for the enamine 5 derived from 1 and phenylacetalde-
hyde. The most stable structure of the iminium ions 3 and 4
was calculated to be (EC,N, EC,C)-s-trans-sc-exo, and the most
stable structure of the enamine 5 was calculated to be the (E)-
s-trans-ap form. For the structure of the pyrrolidine ring, our
calculations suggest that the down conformation tends to be
energetically more favorable than the up conformation
(Figure 2). The energetic preference for down over up confor-
mations is also supported in a recent report by Gschwind et al.
for prolinol and prolinol ether enamines.[11d] There was no elec-
tronic interaction between the silyl atom and the p orbital of
This work was supported by the Mitsubishi Foundation and
a Grant-in-Aid for Scientific Research on Innovative Areas “Ad-
vanced Molecular Transformations by Organocatalysts” from
the Ministry of Education, Culture, Sports, Science, and Tech-
nology, Japan. We thank Tsukuba Advanced Computing Center
for the provision of the computational facilities.
Keywords: asymmetric synthesis
calculations organocatalysis
substituent effects
·
density functional
·
·
reaction mechanisms ·
[1] Selected reviews on organocatalysis: a) Enantioselective Organocatalysis
(Ed.: P. I. Dalko), Wiley-VCH, Weinheim, 2007; b) S. Mukherjee, J. W.
Yang, S. Hoffmann, B. List, Chem. Rev. 2007, 107, 5471; c) A. M. Walji,
D. W. C. MacMillan, Synlett 2007, 1477; d) D. W. C. MacMillan, Nature
2008, 455, 304; e) C. F. Barbas III, Angew. Chem. 2008, 120, 44; Angew.
Chem. Int. Ed. 2008, 47, 42; f) A. Dondoni, A. Massi, Angew. Chem. 2008,
Chem. Eur. J. 2014, 20, 1 – 13
www.chemeurj.org
11
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
120, 4716; Angew. Chem. Int. Ed. 2008, 47, 4638; g) P. Melchiorre, M.
Marigo, A. Carlone, G. Bartoli, Angew. Chem. 2008, 120, 6232; Angew.
Chem. Int. Ed. 2008, 47, 6138; h) S. Bertelsen, K. A. Jørgensen, Chem.
Soc. Rev. 2009, 38, 2178; i) E. Marques-Lopez, R. P. Herrera, M. Christ-
mann, Nat. Prod. Rep. 2010, 27, 1138; j) Science of Synthesis Asymmetric
Organocatalysis 1; Lewis Base and Acid Catalysts (Ed.: B. List), Georg
Thieme Verlag KG, Stuttgart, 2012; k) Science of Synthesis Asymmetric
Organocatalysis 2; Brønsted Base and Acid Catalysts, and Additional
Topics (Ed.: K. Maruoka), Georg Thieme Verlag KG, Stuttgart, 2012.
[2] Review: a) C. Palomo, A. Mielgo, Angew. Chem. 2006, 118, 8042; Angew.
Chem. Int. Ed. 2006, 45, 7876; b) A. Mielgo, C. Palomo, Chem. Asian J.
2008, 3, 922; c) L.-W. Xu, L. Li, Z.-H. Shi, Adv. Synth. Catal. 2010, 352,
243; d) K. L. Jensen, G. Dickmeiss, H. Jiang, L. Albrecht, K. A. Jørgensen,
Acc. Chem. Res. 2012, 45, 248; e) H. Gotoh, Y. Hayashi, Diarylprolinol Silyl
Ethers, Development and Application as Organocatalysts, in Sustainable
Catalysis; (Eds.: P. J. Dunn; K. K. Hii, M. J. Krische, M. T. Williams), John
Wiley & Sons, Hoboken, 2013, pp. 287–316.
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, ꢆ. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[21] The same conformational trend for enamines of diphenylprolinol silyl
ethers has been reported by Seebach and co-workers. D. Seebach, X.
Sun, M.-O. Ebert, W. B. Schweizer, N. Purkayastha, A. K. Beck, J. Dusch-
male, H. Wennemers, T. Mukaiyama, M. Benohoud, Y. Hayashi, M. Reiher,
Helv. Chim. Acta 2013, 96, 799.
[3] Y. Hayashi, H. Gotoh, T. Hayashi, M. Shoji, Angew. Chem. 2005, 117,
4284; Angew. Chem. Int. Ed. 2005, 44, 4212.
[22] L. E. Zimmer, C. Sparr, R. Gilmour, Angew. Chem. 2011, 123, 12062;
Angew. Chem. Int. Ed. 2011, 50, 11860.
[4] M. Marigo, T. C. Wabnitz, D. Fielenbach, K. A. Jørgensen, Angew. Chem.
2005, 117, 804; Angew. Chem. Int. Ed. 2005, 44, 794.
[5] Review, see; A. Erkkilꢂ, I. Majander, P. M. Pihko, Chem. Rev. 2007, 107,
5416.
[23] a) J. L. G. Ruano, V. Marcos, J. Aleman, Chem. Commun. 2009, 4435;
b) A.-N. Alba, X. Companyo, A. Moyano, R. Rios, Chem. Eur. J. 2009, 15,
11095; c) A. Landa, A. Puente, J. I. Santos, S. Vera, M. Oiarbide, C.
Palomo, Chem. Eur. J. 2009, 15, 11954.
[6] a) H. Ishikawa, T. Suzuki, Y. Hayashi, Angew. Chem. 2009, 121, 1330;
Angew. Chem. Int. Ed. 2009, 48, 1304; b) H. Ishikawa, T. Suzuki, H. Orita,
T. Uchimaru, Y. Hayashi, Chem. Eur. J. 2010, 16, 12616; c) H. Ishikawa,
B. P. Bondzic, Y. Hayashi, Eur. J. Org. Chem. 2011, 6020; d) T. Mukaiyama,
H. Ishikawa, H. Koshino, Y. Hayashi, Chem. Eur. J. 2013, 19, 17789; for Ta-
miflu syntheses of other groups using a similar Michael reaction as
a key step, see: e) S. Zhu, S. Yu, Y. Wang, D. Ma, Angew. Chem. 2010,
122, 4760; Angew. Chem. Int. Ed. 2010, 49, 4656; f) J. Rehak, M. Hut’ka,
A. Latika, H. Brath, A. Almassy, V. Hajzer, J. Durmis, S. Toma, R. Sebesta,
Synthesis 2012, 44, 2424; g) J. Weng, Y.-B. Li, R.-B. Wang, G. Lu, Chem-
CatChem 2012, 4, 1007.
[7] H. Ishikawa, M. Honma, Y. Hayashi, Angew. Chem. 2011, 123, 2700;
Angew. Chem. Int. Ed. 2011, 50, 2650.
[8] Y. Hayashi, S. Umemiya, Angew. Chem. 2013, 125, 3534; Angew. Chem.
Int. Ed. 2013, 52, 3450.
[9] Y. Hayashi, H. Gotoh, R. Masui, H. Ishikawa, Angew. Chem. 2008, 120,
4076; Angew. Chem. Int. Ed. 2008, 47, 4012.
[24] E. Gꢅmez-Bengoa, A. Landa, A. Lizarraga, A. Mielgo, M. Oiarbide, C.
Palomo, Chem. Sci. 2011, 2, 353.
[25] Y. Hayashi, M. Toyoshima, H. Gotoh, H. Ishikawa, Org. Lett. 2009, 11, 45.
[26] a) H. Gotoh, R. Masui, H. Ogino, M. Shoji, Y. Hayashi, Angew. Chem.
2006, 118, 7007; Angew. Chem. Int. Ed. 2006, 45, 6853; b) H. Gotoh, H.
Ogino, H. Ishikawa, Y. Hayashi, Tetrahedron 2010, 66, 4894.
[27] H. Gotoh, H. Ishikawa, Y. Hayashi, Org. Lett. 2007, 9, 5307.
[28] a) H. Gotoh, Y. Hayashi, Org. Lett. 2007, 9, 2859; b) Y. Hayashi, S. Saman-
ta, H. Gotoh, H. Ishikawa, Angew. Chem. 2008, 120, 6736; Angew. Chem.
Int. Ed. 2008, 47, 6634.
[29] Recently, our group found that the reaction does not proceed mecha-
nistically through a simple Michael reaction, but involves cyclobutane
and dihydrooxazine-N-oxide intermediates. See ref. [21] and a) K.
Patora-Komisarska, M. Benohoud, H. Ishikawa, D. Seebach, Y. Hayashi,
Helv. Chim. Acta 2011, 94, 719; b) D. Seebach, X. Sun, C. Sparr, M.-O.
Ebert, W. B. Schweizer, A. K. Beck, Helv. Chim. Acta 2012, 95, 1064. Relat-
ed studies by other groups: c) J. Burꢃs, A. Armstrong, D. G. Blackmond,
J. Am. Chem. Soc. 2011, 133, 8822; d) J. Burꢃs, A. Armstrong, D. G. Black-
mond, J. Am. Chem. Soc. 2011, 133, 14264; e) J. Bures, A. Armstrong,
D. G. Blackmond, J. Am. Chem. Soc. 2012, 134, 6741; f) J. Burꢃs, A. Arm-
strong, D. G. Blackmond, Pure Appl. Chem. 2013, 85, 1919; g) G. Sahoo,
H. Rahaman, ꢇ. Madarꢄsz, I. Pꢄpai, M. Melarto, A. Valkonen, P. M. Pihko,
Angew. Chem. 2012, 124, 13321; Angew. Chem. Int. Ed. 2012, 51, 13144;
h) F. Bꢂchle, J. Duschmale, C. Ebner, A. Pfaltz, H. Wennemers, Angew.
Chem. 2013, 125, 12851; Angew. Chem. Int. Ed. 2013, 52, 12619.
[30] G.-L. Zhao, Y. Xu, H. Sunden, L. Eriksson, M. Sayah, A. Cordova, Chem.
Commun. 2007, 734.
[31] Y. Hayashi, H. Gotoh, M. Honma, K. Sankar, I. Kumar, H. Ishikawa, K.
Konno, H. Yui, S. Tsuzuki, T. Uchimaru, J. Am. Chem. Soc. 2011, 133,
20175.
[32] a) Y. Chi, S. M. Gellman, Org. Lett. 2005, 7, 4253. See also the stoichio-
metric Michael additions of enamines derived from prolinol methyl
ether to nitroolefins: b) S. J. Blarer, W. B. Schweizer, D. Seebach, Helv.
Chim. Acta 1982, 65, 1637; c) S. J. Blarer, D. Seebach, Chem. Ber. 1983,
116, 3086.
ˇ
[10] a) D. Seebach, U. Groselj, D. M. Badine, W. B. Schweizer, A. K. Beck, Helv.
ˇ
Chim. Acta 2008, 91, 1999; b) U. Groselj, D. Seebach, D. M. Badine, W. B.
Schweizer, A. K. Beck, I. Krossing, P. Klose, Y. Hayashi, T. Uchimaru, Helv.
ˇ
Chim. Acta 2009, 92, 1225; c) D. Seebach, R. Gilmour, U. Groselj, G.
ˆ
ˆ
Deniau, C. Sparr, M.-O. Ebert, A. K. Beck, L. B. McCusker, D. Sisak, T. Uchi-
maru, Helv. Chim. Acta 2010, 93, 603.
[11] a) M. Cai, Y. Huang, J. Liu, R. Krishnamoorthi, J. Biomol. NMR 1995, 6,
123; b) Y. K. Kang, J. Phys. Chem. B 2007, 111, 10550; c) A. Flores-Ortega,
A. I. Jimꢃnez, C. Cativiela, R. Nussinov, C. Alemꢄn, J. Casanovas, J. Org.
Chem. 2008, 73, 3418; d) M. B. Schmid, K. Zeitler, R. M. Gschwind, Chem.
Sci. 2011, 2, 1793.
[12] R. Liu, J. Zhang, Chem. Eur. J. 2013, 19, 7319.
[13] E. K. Kemppainen, G. Sahoo, A. Piisola, A. Hamza, B. Kꢅtai, I. Pꢄpai, P. M.
Pihko, Chem. Eur. J. 2014, 20, 5983.
[14] M. Silvi, C. Cassani, A. Moran, P. Melchiorre, Helv. Chim. Acta 2012, 95,
1985.
[15] CONFLEX 7, H. Goto, S. Obata, N. Nakayama, K. Ohta, CONFLEX Corpo-
ration, Tokyo, Japan, 2012.
[16] a) T. A. Halgren, J. Comput. Chem. 1999, 20, 720; b) T. A. Halgren, J.
Comput. Chem. 1999, 20, 730.
[17] a) A. D. Becke, Phys. Rev. A 1988, 38, 3098; b) A. D. Becke, J. Chem. Phys.
1993, 98, 1372; c) A. D. Becke, J. Chem. Phys. 1993, 98, 5648; d) C. Lee,
W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.
[33] H. Gotoh, Y. Hayashi, Chem. Commun. 2009, 3083.
[34] a) Z.-J. Jia, H. Jiang, J.-L. Li, B. Gschwend, Q.-Z. Li, X. Yin, J. Grouleff, Y.-C.
Chen, K. A. Jørgensen, J. Am. Chem. Soc. 2011, 133, 5053; b) K. S. Hal-
skov, B. S. Donslund, S. Barfusser, K. A. Jørgensen, Angew. Chem. 2014,
126, 4221; Angew. Chem. Int. Ed. 2014, 53, 4137.
[18] A. P. Scott, L. Radom, J. Phys. Chem. 1996, 100, 16502.
[19] Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215.
[20] Gaussian 09, Revision C.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
Received: May 13, 2014
Published online on && &&, 0000
&
&
Chem. Eur. J. 2014, 20, 1 – 13
www.chemeurj.org
12
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
FULL PAPER
&
Organocatalysis
Y. Hayashi,* D. Okamura, T. Yamazaki,
Y. Ameda, H. Gotoh, S. Tsuzuki,
T. Uchimaru,* D. Seebach
&& – &&
Typecast: Reactions catalyzed by diphe-
nylprolinol silyl ether can be categorized
into three types (see figure): two involve
an iminium ion intermediate, such as
for a Michael-type reaction (type A) and
a cycloaddition reaction (type B), and
one proceeds via an enamine intermedi-
ate (type C). In type A, good enantiose-
lectivity is realized if a catalyst with
a bulky silyl moiety is employed. In
types B and C, good enantioselectivity is
obtained even when the silyl group is
less bulky.
A Theoretical and Experimental Study
of the Effects of Silyl Substituents in
Enantioselective Reactions Catalyzed
by Diphenylprolinol Silyl Ether
Chem. Eur. J. 2014, 20, 1 – 13
www.chemeurj.org
13
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ