First asymmetric cascade reaction catalysed by chiral primary aminoalcohols

Article abstract of DOI:10.1039/c1ob05400g

Readily available chiral primary 1,2-aminoalcohols and diamines have been explored as organocatalysts for a domino Michael-aldol reaction. Their application in this organocascade process afforded cyclohexanone A with high levels of reactivity (up to 91% y

Full text of DOI:10.1039/c1ob05400g

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 5079
www.rsc.org/obc
PAPER
First asymmetric cascade reaction catalysed by chiral primary
aminoalcohols†‡
Carlos Arro´niz,a Carmen Escolano,*a F. Javier Luque,*b Joan Boscha and Mercedes Amata
Received 14th March 2011, Accepted 11th April 2011
DOI: 10.1039/c1ob05400g
Readily available chiral primary 1,2-aminoalcohols and diamines have been explored as organocatalysts
for a domino Michael–aldol reaction. Their application in this organocascade process afforded
cyclohexanone A with high levels of reactivity (up to 91% yield) and stereoselectivity (>97 : 3 d.r., up to
93% ee). Depending on the acid cocatalyst different chiral species (cyclic secondary amines vs. acyclic
primary amines) might catalyse the process. In order to shed light on the catalytic activation, several
experiments were carried out and a detailed possible reaction mechanism is proposed. Theoretical
studies support the stereochemical outcome of the process.
In this context, chiral primary 1,2-aminoalcohols, easily ac-
Introduction
cessible from inexpensive natural amino acids, are valuable small
molecules that have been efficiently used as chiral auxiliaries in the
preparation of enantiomeric scaffolds giving access to structurally
complex bioactive and natural products.7 Furthermore, a large
number of vicinal amino alcohol derivatives have successfully
acted as ligands in a great variety of transition metal catalytic
reactions.8 Although both these applications are well documented,
chiral primary 1,2-aminoalcohols have been scarcely studied as
“metal-free catalysts”9 and, to the best of our knowledge, their
application in asymmetric cascade reactions remains unexplored.
Going a step further in the study of chiral aminoalcohols
in asymmetric synthesis, we herein examine their usefulness as
organocatalysts for the enantioselective production of compounds
by a cascade reaction. The structural requirements of the catalyst
for efficient stereoselective catalysis are discussed. Finally, compu-
tational studies are performed to identify clues about the origin of
the stereoselectivity.
In the last decade, organocatalysis has emerged as one of the
most exciting and rapidly expanding fields in organic chemistry.1
In particular, aminocatalysis has provided the life sciences with
easy and environmentally friendly access to important building
blocks.2 Secondary amines have played a key role within this field
due to their efficient activation of carbonyl compounds such as
aldehydes, a,b-unsaturated aldehydes and ketones. Moreover, the
past few years have witnessed the rapid development of chiral pri-
mary amine-based organocatalysts3 owing to their extraordinary
usefulness and versatility as activators of a,b-unsaturated ketones
in asymmetric catalysis.4
In parallel with the advances in the use of primary
amine organocatalysts, significant developments in asymmetric
organocascade reactions have also been reported.5 It is not
surprising that domino processes represent a flourishing area
in organic chemistry since they allow a number of bonds and
stereocenters to be efficiently created in a single operation. These
straightforward routes to molecular complexity are important in
green chemistry for their atom economy, reduction of synthetic
steps, and minimization of solvents and waste.6 It is thus highly
desirable to provide readily available molecules able to promote
asymmetric cascade reactions.
Results and discussion
In a seminal study by Jørgensen’s group, a phenylalanine-derived
imidazolidine 1 was shown to catalyze a highly enantio- and
diastereoselective domino Michael–aldol reaction of acyclic b-
ketoesters and a,b-unsaturated ketones to form optically active
cyclohexanones (Scheme 1).10
An initial screening was carried out to optimize the reaction
conditions of benzylideneacetone with benzyl benzoylacetate
using 10 mol% of 1. The highest reaction rates were observed
in EtOH at room temperature for 44 h (89% yield), with excellent
diastereoselectivity (>97 : 3 d.r.) and good enantiomeric excess
(89% ee). This transformation was also performed at 10 ◦C (80%
yield, >97 : 3 d.r., 95% ee), although the reaction time increased to
190 h. The reaction also proceeded under solvent-free conditions,
even though the yield was significantly lower (60%). Notably,
aLaboratory of Organic Chemistry, Faculty of Pharmacy, and Institute of
Biomedicine (IBUB), 08028 Barcelona, Spain. E-mail: cescolano@ub.edu
bDepartment of Physical Chemistry, Faculty of Pharmacy, University of
Barcelona, Av. Diagonal 643, 08028 Barcelona, Spain. E-mail: fjluque@
ub.edu
† Dedicated to the memory of Prof. Rafael Suau.
‡ Electronic supplementary information (ESI) available: Copies of 1H and
13C NMR spectra of new products 1 and 13, copies of the HPLC chro-
matograms, representation of alternative conformations for the products
with R and S configuration, and optimized geometries of the structures
shown in Fig. 4, 5, S1 and S2. See DOI: 10.1039/c1ob05400g
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 5079–5085 | 5079
©

View Article Online
Thus, when we directly used diamine 3 in the presence of an
equimolecular amount of glyoxylic acid (Table 1, entry 2), we
obtained A with the same optical purity (89% ee) as when using
preformed imidazolidine 1 (vide supra). In order to elucidate the
existing species in the reaction media, equimolecular amounts of
diamine 3 and glyoxylic acid were mixed in EtOH, the solvent of
the domino reaction. Under these conditions, total conversion to
imidazolidine 1 was observed after 1 h.15 Additionally, the mixture
of equimolecular amounts of both reagents, 3 and glyoxylic acid,
in CD3OD in an NMR tube allowed us to confirm that the
formation of 1 was instantaneous. These findings indicate that
the phenylalanine-derived imidazolidine 1 is likely to be present
in the domino reaction media (EtOH). Similarly, oxazolidine 2
is also rapidly formed in EtOH from glyoxylic acid monohydrate
and (S)-phenylalaninol 4. Thus, the previous preparation of cyclic
catalysts 1 and 2 is not required since in situ formation can occur
in the reaction media.
With this information in hand, we turned our attention to the
study of primary aminoalcohols as organocatalysts, carrying out
a screening of the reaction conditions. The use of 20 mol% of (S)-
phenylalaninol (4) and glyoxylic acid furnished cyclohexanone
A in 77% yield and 73% ee (Table 1, entry 3). Interestingly, an
optimal ratio of the aminoalcohol and the acid (20 and 10 mol%,
respectively) significantly increased the yield and enantioselectivity
of the process (Table 1, entry 4). This result prompted us to explore
the role of the acid in the reaction.
Scheme 1 Organocatalytic asymmetric domino Michael–aldol reaction.
a single enantiomer (A) precipitated from the reaction mixture,
which was isolated in a pure form simply by filtration, suggest-
ing that a continuous recrystallization took place during the
process.
At the outset of our studies, we planned to explore the
influence of an oxygen atom within the cyclic catalyst. Thus,
4-benzyloxazolidine-2-carboxylic acid 2, an analogue of the
Jørgensen imidazolidine 1, was used as an organocatalyst in the
model reaction.11 Satisfactorily, novel oxazolidine 2 was indeed
able to catalyse the reaction (Table 1, entry 1), thus indicating
its capacity to activate a,b-unsaturated ketones.12 In this context,
it is known that cyclic secondary amines are reluctant to react
with a,b-unsaturated ketones due to the formation of congested
iminium adducts, and that primary amines can easily overcome
such steric difficulties.4 Interestingly, some authors consider that
one of the many imidazolidines that catalyse 1,4-additions is in fact
a precatalyst,13 since the primary amine derived from its hydrolysis
can catalyse the conjugate addition.14
With the results obtained so far we assumed that the catalyst of
the reaction could be either the heterocycles (1 or 2) or the primary
amines (3 or 4), with the glyoxylic acid acting as a cocatalyst. To
discard the former possibility, glyoxylic acid was substituted by
other acids that prevent the in situ formation of the five-membered
ring derivatives 1 and 2 from diamine 3 and aminoalcohol 4,
respectively.
Bearing this case in mind, additional experiments were envis-
aged considering that imidazolidine 1 and oxazolidine 2 may be in
equilibrium with the diamine 3 and aminoalcohol 4, respectively.
The use of benzoic acid as the cocatalyst in the presence of
diamine 3, gave the product A in 82% yield and 93% ee (Table 1,
entry 5). Aminoalcohol 4 was also able to catalyse the reaction in
the presence of benzoic acid (Table 1, entry 6). These experiments
proved that primary amines 3 and 4 are able to catalyze the domino
Michael–aldol reaction on their own. However, it can not be
completely excluded that reactions performed with glyoxylic acid
(Table 1, entries 2–4) are not catalyzed by heterocycle 1 or 2.
The product A was also obtained by using AcOH as the
cocatalyst, although the yield and the ee decreased to 60 and 53%,
respectively (Table 1, entry 7). Other acids with lower pKa such
as oxalic and trifluoroacetic acids turned out to be ineffective and
the reaction did not progress at all.
Table 1 Optimization of the diastereo- and enantioselective domino
Michael–aldol reaction by 2–4a
Catalyst
Entry [mol%] Acid [mol%]
Solvent Time/d Yield (%)b ee (%)c
1
2
3
4
5
6
7
8
2 [20]
3 [10]
4 [20]
4 [20]
3 [10]
4 [20]
4 [20]
4 [20]
4 [20]
4 [20]
4 [20]
4 [20]
EtOH
Glyoxylic [10] EtOH
Glyoxylic [20] EtOH
Glyoxylic [10] EtOH
7
3
6
6
4
6
6
5
6
6
2
3
65
70
77
88
82
66
60
42
70
82
90
91
70
89
73
82
93
66
53
86
86
81
75d
78d
In order to provide a more environmentally friendly procedure,
water and neat conditions were considered. Although previous
studies using 1 in water were not successful,10 (S)-phenylalaninol
(4) enabled the reaction to take place in water with high selectivity
but moderate reactivity (Table 1, entry 8), the latter probably due
to the low solubility of the reagents in the aqueous media.
Neat conditions in the presence of glyoxylic acid resulted in an
improved enantioselectivity (Table 1, entry 9 vs. entry 3), whereas
with benzoic acid, the absence of solvent enhanced both the
reaction rate and enantioselectivity (Table 1, entry 10 vs. entry
6).
Benzoic [10]
Benzoic [20]
Acetic [20]
EtOH
EtOH
EtOH
Glyoxylic [20] H2O
Glyoxylic [20] neat
9
10
11
12
Benzoic [20]
Glyoxylic [20] neat
Benzoic [20] neat
neat
All the reactions were performed on a 0.5 mmol scale at room temperature
unless indicated otherwise (see Experimental section).a Diastereoisomeric
ratio in all reactions showed to be >97 : 3 determined by 1H NMR spec-
troscopy of the crude product. b Yields of isolated product. c Determined
by chiral stationary phase HPLC. d Reaction performed at 40 ◦C.
In the domino reaction, heating the mixtures at 40 ◦C in solvent-
free conditions reduced the reaction times and increased the yields
5080 | Org. Biomol. Chem., 2011, 9, 5079–5085
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
(>90%), but slightly decreased enantioselectivities (Table 1, entries
An additional stereocenter with a tethered phenyl group was in-
cluded in the catalyst, and (1S,2R)-2-amino-1,2-diphenylethanol
7 was examined with either glyoxylic or benzoic acid as the
cocatalyst. Such conditions afforded ent-A in moderate yield
and enantioselectivity (Table 2, entries 3 and 4). Running the
reaction in the presence of its 1R, 2R diastereoisomer 8 did not
significantly improve the rate or selectivity of the process (Table
2, entry 5). These results indicated that only the stereocenter
bearing the primary amino group has a significant influence on
the stereochemical outcome of the reaction. These observations
were confirmed by the different behavior of the constrained
diastereoisomers 9 and 10. Using trans-(R,R)-aminoindanol 9,
ent-A was formed in 65% yield and 70% ee, whereas with cis-(S,R)-
aminoindanol 10, low conversion to the opposite enantiomer (A)
occurred, albeit with a higher ee value (Table 2, entries 6 and 7).
When bearing a 3-indolyl substituent the primary aminoalcohol
catalyst [(S)-tryptophanol, 11] was ineffective (Table 2, entry 8).
Electron-donating groups in the phenyl substituent hampered
the reaction evolution as (S)-3,4-dimethoxyphenylalaninol (12)
needed a higher temperature in solvent-free conditions to yield
50% of A (Table 2, entry 9).
11 and 12).16
Other vicinal diamine and aminoalcohols as domino Michael–aldol
reaction catalysts
Since diamine 3 and (S)-phenylalaninol (4) are good candidates
for the model domino reaction, other 1,2-aminoalcohol derivatives
with different structural features were studied (Fig. 1). Represen-
tative results are summarized in Table 2.
Fig. 1 Diamine 5 and aminoalcohols 6–13 used in this study. TMS =
trimethylsilyl.
The above results show that, in the aminoalcohol series, the
optimal aromatic substituent for an efficient domino Michael–
aldol reaction is a phenyl or benzyl group.
Alcohol protection of (S)-phenylalaninol (4) gave trimethylsilyl
ether derivative 13, which catalyzed the cascade reaction in good
yield. Nevertheless, the absence of a hydrogen-bonding donor
decreased the reaction enantioselectivity (Table 2, entry 10).
We first examined the influence of the aromatic ring arrange-
ment within the catalyst. Thus, when the aromatic substituent
was directly linked to the chiral stereocenter of the diamine (5) the
effect was considerable, with the ee dropping from 89 to 68% (Table
1, entry 2 vs. Table 2, entry 1). However, such a phenyl substituent
within an aminoalcohol moiety (6) affected the reaction outcome
only slightly (Table 2, entry 2 vs. Table 1, entry 3), this compound
thereby showing more tolerance to structural modifications.
As expected, the use of (R)-phenylglycinol (6) as the catalyst led
to the enantiomer (ent-A) of the final product obtained in all the
aforementioned studies using compounds 3–5 as the chiral source
(Table 2, entry 3).
Mechanistic and theoretical considerations
Based on our experimental observations and previous
considerations,10 a putative mechanism can be proposed for the
Michael–aldol reaction.
Two different pathways may be considered depending on the
catalytic participation of cyclic secondary amine IA (cycle A
depicted in Fig. 2) or acyclic primary amine IB (cycle B). Using
catalysts 3 or 4 with glyoxylic acid as the cocatalyst, the reaction
can progress through both primary amine catalysis17 (cycle B)
and secondary amine catalysis18 (cycle A). However, when the
reaction was performed in the presence of other acid cocatalysts
such as benzoic acid, the reaction course would only follow cycle
B (X = NCH3, O; R1 = Bn; R2 = Ph) since the absence of aldehyde
functionality within the acid would avoid the cyclic secondary
amine formation.
The iminium/enamine cascade sequence would proceed
through the initial activation of the unsaturated ketone by either
the catalyst IA or IB to form iminium ions IIA and IIB, respectively.
The ketoester nucleophilic attack at the Si face of these activated
eniminium ions would afford Michael adducts IIIA and IIIB,
with two new stereogenic centers, although one of them can
probably epimerize in the reaction media. After an enamine
isomerization, intermediates IVA and IVB can adopt a prefer-
ential pseudo-chair conformation where the bulky substituents
are in equatorial disposition. Only these favourable conformers,
probably assisted by a hydrogen-bonding interaction, would easily
undergo the aldol reaction. Finally, subsequent hydrolysis of
VA and VB would afford cyclohexanone A and re-establish the
catalysts.
Table 2 Domino Michael–aldol reaction catalysed by diamine 5 and
aminoalcohols 6–13a
Entry Catalyst Acid
Solvent Time/d Yield (%)b ee (%)c
1
2
3
4
5
6
7
8
9
5
Glyoxylic EtOH
Glyoxylic EtOH
Glyoxylic EtOH
3
6
6
4
6
6
6
4
3
6
70
71
67
55
55
65
28
42
50
77
68d
-80
-73
-63
-77
-70
86
6
7
7
Benzoic
EtOH
8
Glyoxylic EtOH
Glyoxylic EtOH
Glyoxylic EtOH
Glyoxylic EtOH
Glyoxylic neat
Glyoxylic EtOH
9
10
11
12
13
56
79e
61
10
All the reactions were performed on a 0.5 mmol scale at room temperature
with 20 mol% of the cocatalyst and catalysts unless indicated otherwise (see
Experimental section).a Diastereoisomeric ratio in all reactions showed to
be >97 : 3 determined by 1H NMR spectroscopy of the crude product.
b Yields of isolated product. c Determined by chiral stationary phase
HPLC. d Reaction performed with 10 mol% of both diamine 5 and glyoxylic
acid. e Reaction performed at 40 ◦C.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 5079–5085 | 5081
©

View Article Online
Fig. 2 Proposed Michael–aldol reaction mechanism catalysed by primary and secondary amines.
Interactions between catalysts and other carbonyl groups
present in the mixture may produce reversible side reactions. How-
ever, these would not interfere with the organocascade reaction as
the whole system is in equilibrium wherein the precipitation of
the final product (A) from the reaction mixture acts as the driving
force of the process.
Regarding the stereoselectivity in the domino Michael–aldol
sequence, the key step for the asymmetric induction is the initial
Michael addition, which involves the generation of the two first
stereogenic centers. Since only one of them is configurationally
stable, this one determines the configuration of other stereocenters
present in the final product (A). Thus, we centered our attention
on the stereochemical outcome during the formation of this
stable stereocenter. To explain the stereochemistry of the Michael
adduct IIIB, two main factors were considered: the stability of the
iminium ion (IIB) and the stereofacial selectivity in the ketoester
attack.
The interaction of the catalyst with the a,b-unsaturated ketone
could originate four possible iminium derivatives (Fig. 3). The
transoid iminium salts were expected to be more stable due to the
larger steric hindrance in the cisoid derivatives. Indeed, B3LYP/6-
31G(d) calculations (phenylglycinol chosen as model, Fig. 3 X =
O) indicate that transoid forms were more stable by 2.4 kcal mol-1.
Among them, trans-im2 was more stable than trans-im1 by 0.7 kcal
mol-1. This effect can be ascribed to the balance of two opposite
effects (Fig. 4): (i) the energetic stabilization due to the formation
of an intramolecular hydrogen bond between the hydroxyl group
and the iminium N–H, and (ii) the destabilization due to the
proximity of the benzene ring to the carbon atom in the position
b to the iminium nitrogen and trans to the iminium hydrogen
(CH in trans-im2; CH3 in trans-im1), which is larger in trans-im1
due to the presence of the methyl group. Relief of such steric
hindrance, which is associated with the loss of the intramolecular
hydrogen bond, does not afford a net gain in stability, as the corre-
sponding conformations are destabilized by around 1.1 kcal mol-1
(Fig. 4).
Fig. 3 Putative iminium salt derivatives and Michael adduct intermediate.
The nucleophilic attack of benzyl benzoylacetate at the Si
face of the iminium intermediate (trans-im2) will determine an
R configuration in the newly created stereocenter [(R)-IIIB where
R1 = Ph, Fig. 3]. In order to gain insight into the preference for
the Si face, we determined the relative stability of the products
formed from the attack of benzyl benzoylacetate at the two faces
of trans-im2. Exploration of the different conformational states
for the products with R (Si face, (R)-IIIB) and S (Re face, (S)-
IIIB) configurations indicated that the former is predicted to be
more stable by 2.0 kcal mol-1 at the B3LYP/6-31G(d) level (Fig.
5). In the two cases an intramolecular hydrogen bond is formed
between the hydroxyl unit and the carbonyl group of the keto
or ester moieties upon attack via Si or Re faces, respectively (the
alternative conformations where hydrogen bonding involves the
ester (Si face) or keto (Re face) units are destabilized by 2.9 and
3.6 kcal mol-1; see Fig. S1 in the ESI†). Inspection of Fig. 5
suggests that the preference for the product with R configuration
5082 | Org. Biomol. Chem., 2011, 9, 5079–5085
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
benzyl ring and the hydrogen atom at the chiral center in the S
product.
On the basis of these findings, it can be hypothesized that the
formation of the transition state (TS) leading to the R product (Si
face, (R)-IIIB) will be energetically more favorable, which would
thus justify the stereochemical outcome of the reaction. In order
to confirm this hypothesis, the TS leading to the products were
determined and confirmed by the presence of a single imaginary
frequency, which mainly involves the displacement of the carbon
atoms involved in the formation of the bond between trans-
im2 and benzyl benzoylacetate (C–C distance of 2.31 and 2.33
for transition states via Si and Re faces, respectively; see Fig.
S2 in the ESI†). Moreover, the TS yielding the S product is
destabilized by 3.6 kcal mol-1 relative to the TS leading to the
R product [(R)-IIIB]. Notably, such a difference in stability can
be attributed, at least in part, to the stronger intramolecular
hydrogen bond in the R product-TS, as noted in a distance of
2.77 A, which compares with a value of 3.02 A found for the
S product-TS (see Fig. S2 in the ESI†). Overall, these findings
highlight the important role played by the hydroxyl group of
the aminoalcohol catalyst. This observation is supported by our
experimental results, comparing the ee in the reaction using 4
(Table 1, entry 4, ee 82%) and its silylated analogue 13 (Table 2,
entry 10, ee 61%).
Fig. 4 Representation of the most favorable species of trans-im1 and tran-
s-im2, their relative energy (kcal mol-1; in bold), and selected intramolecu-
lar distances (A; in italics) determined from geometry optimizations at the
˚
˚
˚
B3LYP/6-31G(d) level.
Finally, the a-dicarbonylic stereocenter is epimerizable and
its configuration may be determined by the subsequent aldolic
condensation step, taking into account that the most stable
chair-like conformation for IVA and IVB would involve the
pseudoequatorial disposition of all substituents (Fig. 2). In this
last step, a hydrogen-bonding interaction between the hydroxyl
group and the ketone is likely to be present, as indicated in the
theoretical calculations for (R)-IIIB (Fig. 5).
Conclusions
Chiral aminoalcohols and diamines, which are readily available
from naturally occurring amino acids, have proven to be new, use-
ful organocatalysts for a highly stereoselective domino Michael–
aldol reaction, affording cyclohexanone A with excellent yields (up
to 91%), diastereoselectivities (>97 : 3 d.r.) and enantioselectivities
(up to 93% ee). Initial studies using cyclic secondary amines 1 and
2, and acyclic primary amines 3 and 4, in the presence of glyoxylic
acid showed that two different types of catalysis might be involved.
Benzoic acid cocatalysis pointed to primary amine catalysis, in
which 3 and 4 acted as the real catalysts. However, when glyoxylic
acid cocatalysed the reaction, the heterocyclic species present in
the media could take part in the process through secondary amine
catalysis. A detailed mechanism for the organocascade reaction
was proposed.
All aminoalcohol derivatives studied were able to catalyse the
domino process. However, few structural catalyst modifications
allow an efficient reaction, thereby indicating that this transfor-
mation is very sensitive to the catalyst structure.
Finally, theoretical calculations have been performed to justify
the stereochemical outcome of the Michael addition, which
determines the configuration of the three contiguous stereogenic
centres present in the final product. From these studies we can
conclude that the preference for the product with R configuration
(attack via Si face) can be attributed to the subtle balance
Fig. 5 Two views of the most stable products formed upon nucleophilic
attack of the benzyl benzoylacetate (shown in yellow) to the Si/Re faces
of the iminium intermediate trans-im2 (shown in blue). The relative energy
(kcal mol ; in bold), and selected intramolecular distances (A; in italics)
determined from geometry optimizations at the B3LYP/6-31G(d) level are
shown.
-1
˚
arises from two factors: (i) a stronger hydrogen bond between
hydroxyl and carbonyl groups in the R product, as noted in the
˚
shorter O ◊ ◊ ◊ H distance (2.95 and 3.08 A for attack via Si or
Re faces, respectively), and (ii) a larger steric clash between the
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 5079–5085 | 5083
©

View Article Online
1
between (i) the hydrogen bonding between the hydroxyl group
of the transoid iminium intermediate and the carbonyl unit of the
benzyl benzoylacetate reagent, and (ii) the steric clash between the
reacting partners.
a pale yellow solid (2.52 g, 85%). H NMR (400 MHz, CDCl3)
d 0.00 (s, 9H, 3CH3), 1.30 (br s, 2H, NH2), 2.38 (dd, J = 13.4,
8.4 Hz, 1H, CH2Ph), 2.66 (dd, J = 13.4, 5.2 Hz, 1H, CH2Ph), 2.99
(m, 1H, CH), 3.28 (dd, J = 9.8, 6.8 Hz, 1H, CH2O), 3.44 (dd,
J = 9.8, 4.3 Hz, 1H, CH2O), 7.05–7.21 (m, 5H, PhH); 13C NMR
(100 MHz, CDCl3) d -0.6 (CH3), 40.4 (CH2Ph), 54.1 (CH), 67.2
(CH2O), 126.2 (Ph), 128.4 (Ph), 129.2 (Ph), 139.1 (C-i); [a]2D2 -8.5
(c 1.0, CHCl3); HMRS C12H22NOSi (M + H+), 224.1465; found,
224.1468.
Experimental
Unless otherwise indicated, NMR spectra were recorded in CDCl3
at 400 MHz (1H) and 100.6 MHz (13C), and chemical shifts are
reported in d values downfield from TMS or relative to residual
chloroform (7.26 ppm, 77.0 ppm) as an internal standard. Data
are reported in the following manner: chemical shift, multiplicity,
coupling constant (J) in hertz (Hz), integrated intensity. Multi-
plicities are reported using the following abbreviations: s, singlet;
dd, doublet of doublets; m, multiplet; br s, broad signal. Optical
rotations were measured on a Perkin–Elmer 241 polarimeter.
The enantiomeric excess (ee) of the products was determined by
chiral stationary phase HPLC (Daicel Chiralpak IA column) with
methyl tert-butyl ether/EtOH 98/2 as eluent. Analytical grade
solvents and commercially available reagents were used without
further purification. Benzyl benzoylacetate,19 diamines 3 and 5,13a
aminoalcohols 4 and 6,20 and 1221 were prepared according to
literature procedures.
General procedure for the domino Michael–aldol reaction
To a solution of benzylideneacetone (0.5 mmol) and benzyl
benzoylacetate (1 mmol) in EtOH (1 mL), the appropriate catalyst
(0.05–0.1 mmol) and acid cocatalyst (if appropriate) were added.
The reaction mixture was stirred at the temperature and for the
time indicated in the tables. The reaction mixture was diluted with
Et2O (2 mL) and filtered. The precipitate was collected to give the
cyclohexanone product as a white solid. The enantiomeric excess
(ee) was determined by HPLC analysis with an IA column (methyl
tert-butyl ether/EtOH 98/2, 1.0 mL min-1, l = 210 nm).
Theoretical computations
Geometrical parameters and energy differences were determined
from full geometry optimizations at the B3LYP/6-31G(d) level.
For the (both cis and trans) iminium derivatives a conformational
exploration was performed around the rotatable N–C bond in
order to determine the minimum energy conformation. Regarding
the products (with both R and S configuration) formed upon
attack of the benzyl benzoylacetate to trans-im2, a systematic
exploration was performed to explore different hydrogen bonding
patterns as well as the orientation of the benzyl and benzoyl
moieties. The minimum energy nature of the stationary points
was verified by inspection of the vibrational frequencies, which
were positive in all cases with the exception of the transition
states, which were characterized by a single imaginary frequency.
Calculations were performed using Gaussian-03.22
(4S)-4-Benzyl-1-methylimidazolidine-2-carboxylic acid (1)
In a NMR tube glyoxylic acid monohydrate (4.6 mg, 0.05 mmol)
was added to a solution of (S)-1-amino-2-methylamino-3-
phenylpropane (3; 8.5 mg, 0.05 mmol) in CD3OD (1 mL). The 1H
NMR spectrum was immediately recorded, indicating the presence
1
of a 1 : 1 mixture of diastereoisomeric imidazolidines. H NMR
(400 MHz, CD3OD) d 2.64–2.75 (m, 2H), 2.82 (s, 6H, NCH3),
2.86–2.94 (m, 2H), 3.03–3.09 (m, 2H), 3.15–3.20 (m, 2H), 3.52
(m, 1H, CH–N, one diastereoisomer), 3.82 (m, 1H, CH–N, one
diastereoisomer), 4.34 (m, 1H. CH–N, one diastereoisomer), 4.38
(m, 1H, CH–N, one diastereoisomer), 7.23–7.33 (m, 10H, ArH);
13C NMR (100 MHz, CD3OD) d 39.1. 39.6, 40.3, 40.7, 59.3, 59.8,
60.0, 60.3, 84.8, 86.6, 128.0, 128.4, 129.6, 129.8, 130.0, 130.1, 130.4,
130,5, 138.6, 138.7, 170.1, 170.6.
Acknowledgements
(4S)-4-Benzyloxazolidine-2-carboxylic acid (2)
Financial support from the Ministry of Science and Innovation,
Spain (Projects CTQ2009-07021/BQU and SAF2008-05595), and
the AGAUR, Generalitat de Catalunya (Grants 2009-SGR-111
and 2009-SGR-249) is gratefully acknowledged. C.A. thanks the
Ministry of Education and Science for a predoctoral fellowship.
The Centre de Supercomputacio´ de Catalunya (CESCA) is
acknowledged for computational facilities.
An equimolecular mixture of (S)-phenylalaninol (151 mg, 1 mmol)
and glyoxylic acid monohydrate (92 mg, 1 mmol) in CH2Cl2
(20 mL) was stirred at room temperature overnight. The solvent
was evaporated to give quantitatively the desired product as a
white solid as a mixture of diastereoisomers. HMRS C11H14NO3
(M + H+), 208.0968; found, 208.0968.
(S)-1-Phenyl-3-(trimethylsilyloxy)propan-2-amine (13)
Notes and references
Iodine (46 mg, 0.18 mmol) and hexamethyldisilazane (2.8 mL,
13.23 mmol) were added to a solution of (S)-phenylalaninol
(2 g, 13.23 mmol) in anhydrous CH2Cl2 (45 mL). The resulting
yellowish mixture was stirred at room temperature for 4 h. The
reaction was quenched with MeOH (1 mL) and the mixture was
evaporated to give a colourless oil that was dissolved in CH2Cl2
(45 mL). A solution of Na2S2O3 (4 g) in water was added, the
mixture was stirred for 5 min, and the phases were separated.
The organic phase was dried, filtered and concentrated to give
1 (a) A. Berkessel, H. Gro¨ger, Asymmetric Organocatalysis, Wiley-VCH,
Weinheim, 2005; (b) P. I Dalko, Enantioselective Organocatalysis,
Wiley-VCH, Weinheim, 2007; (c) H. Pellissier, Recent Developments
in Asymmetric Organocatalysis, RCS Publishing, 2010. For selected
reviews, see: (d) Special Issue on Asymmetric Organocatalysis: Acc.
Chem. Res., 2004, 37, ((8)); (e) Special edition on organocatalysis:
Chem. Rev., 2007, 107; (f) D. W. C. MacMillan, Nature, 2008, 455,
304–308; (g) S. Bertelsen and K. A. Jørgensen, Chem. Soc. Rev.,
2009, 38, 2178–2189. For reviews on organocatalytic synthesis of drugs
and natural products, see: (h) M. J. Gaunt, C. C. C. Johansson, A.
McNally and N. T. Vo, Drug Discov. Today, 2007, 12, 8–27; (i) R. M.
5084 | Org. Biomol. Chem., 2011, 9, 5079–5085
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
De Figueiredo and M. Christmann, Eur. J. Org. Chem., 2007, 2575–
2600; (j) E. Marque´s-Lo´pez, R. P. Herrera and M. Christmann, Nat.
Prod. Rep., 2010, 27, 1138–1167.
2 For selected reviews on asymmetric aminocatalysis, see: (a) B. List,
Chem. Commun., 2006, 819–824; (b) P. Melchiore, M. Marigo, A.
Carlone and G. Bartoli, Angew. Chem., Int. Ed., 2008, 47, 6138–6171.
For an instructive essay on the aminocatalysis evolution, see: (c) B. List,
Angew. Chem., Int. Ed., 2010, 49, 1730–1734.
(d) J. Pulkkinen, P. S. Aburel, N. Halland and K. A. Jørgensen, Adv.
Synth. Catal., 2004, 346, 1077–1080.
14 A mechanism involving a diamine catalyst (diphenylethylenediamine),
instead of the previously supposed imidazolidine catalyst, has been
proposed: (a) H. Kim, C. Yen, P. Preston and J. Chin, Org. Lett., 2006,
8, 5239–5242. For selected organocascade reactions of enones catalysed
by primary amines, see: (b) L.-Y. Wu, G. Bencivenni, M. Mancinelli, A.
Mazzanti, G. Bartoli and P. Melchiorre, Angew. Chem., Int. Ed., 2009,
48, 7196–7199; (c) G. Bencivenni, L.-Y. Wu, A. Mazzanti, B. Giannichi,
F. Pesciaioli, M.-P. Song, G. Bartoli and P. Melchiorre, Angew. Chem.,
Int. Ed., 2009, 48, 7200–7203; (d) Y.-Q. Yang, Z. Chai, H.-F. Wang,
X.-K. Chen, H.-F Cui, C.-W. Zheng, H. Xiao, P. Li and G. Zhao,
Chem.–Eur. J., 2009, 15, 13295–13298; (e) H.-F. Cui, Y.-Q. Yang, Z.
Chai, P. Li, C.-W. Zheng, S.-Z. Zhu and G. Zhao, J. Org. Chem., 2010,
75, 117–122. For an example of a primary amine (phenylethylamine)
catalysed Michael reaction, see: (f) Y. Hirai, T. Terada, T. Yamazaki
and T. Momose, J. Chem. Soc., Perkin Trans. 1, 1992, 509–516.
15 The preparation of 1 was previously reported by reaction of glyoxylic
acid monohydrate and the corresponding diamine 3 in CH2Cl2 for 15
h (see ref. 10).
3 For selected reviews on primary amine catalysis, see: (a) F. Peng and Z.
Shao, J. Mol. Catal. A: Chem., 2008, 285, 1–13; (b) L.-W. Wu and Y.
Lu, Org. Biomol. Chem., 2008, 6, 2047–2053; (c) L.-W. Xu, J. Luo and
Y. L u , Chem. Commun., 2009, 1807–1821. For a review on asymmetric
primary amine catalysts based on Cinchona alkaloids, see: (d) Y.-C.
Chen, Synlett, 2008, 1919–1930.
4 G. Bartoli and P. Melchiorre, Synlett, 2008, 1759–1772.
5 For reviews on organocatalytic domino/cascade reactions, see: (a) D.
Enders, C. Grondal and M. R. M. Hu¨ttl, Angew. Chem., Int. Ed.,
2007, 46, 1570–1581; (b) B. Westermann, M. Ayaz and S. S. van
Berkel, Angew. Chem., Int. Ed., 2010, 49, 846–849; (c) C. Grondal, M.
Jeanty and D. Enders, Nat. Chem., 2010, 2, 167–178. For a review on
asymmetric cascade reactions catalyzed by chiral secondary amines,
see: (d) X. Yu and W. Wang, Org. Biomol. Chem., 2008, 6, 2037–
2046. For a review on asymmetric cascade reactions catalyzed by chiral
amines, see: (e) S. Vera and P. Melchiore, An. Qu´ım., 2010, 106, 277–
284.
16 Performing the reaction at lower temperatures did not result in better
reactivity or selectivity.
17 In the primary amine catalysis, the glyoxylic acid would only act
as an acidic proton source, which would accelerate the iminium salt
formation.
6 (a) K. C. Nicolaou, D. J. Edmonds and P. G. Bulger, Angew. Chem.,
Int. Ed., 2006, 45, 7134–7186; (b) K. C. Nicolaou and J. S. Chen, Chem.
Soc. Rev., 2009, 38, 2993–3009.
18 In the secondary amine catalysis, the cis heterocyclic diastereoisomer
(IA) would play a major role as catalyst. In a related case, a 1,4-cis
heterocyclic diastereoisomer was shown to be more reactive than the
trans isomer in an N-acylation process. This can be rationalized taking
into account that substituents in trans relationship shield the nitrogen
inhibiting a nucleophilic attack. For the related example, see: W. Ando,
Y. Igarashi and L. Huang, Chem. Lett., 1987, 1361–1364See also ref.
11.
19 C.-Q. Sun, P. T. W. Cheng, J. Stevenson, T. Dejneka, B. Brown, T. C.
Wang, J. A. Robl and M. A. Poss, Tetrahedron Lett., 2002, 43, 1161–
1164.
20 S. E. Howson, L. E. N. Allan, N. P. Chmel, G. J. Clarkson, R. Gorkum
and P. Scott, Chem. Commun., 2009, 1727–1729.
7 For a review on 1,2-aminoalcohols as chiral auxiliaries in asymmetric
synthesis, see: (a) D. J. Ager, I. Prakash and D. R. Schaad, Chem.
Rev., 1996, 96, 835–875. For reviews on our work in the field, see:
(b) C. Escolano, M. Amat and J. Bosch, Chem.–Eur. J., 2006, 12, 8198–
8207; (c) M. Amat, M. Pe´rez and J. Bosch, Synlett, 2011, 143–160.
8 For a review on catalytic applications of nitrogen-containing ligands,
including 1,2-aminoalcohols, see: F. Fache, E. Schultz, M. L. Tom-
masino and M. Lemaire, Chem. Rev., 2000, 100, 2159–2232.
9 For examples on the use of chiral primary aminoalcohols as metal-free
organocatalysts, see: (a) A. V. Malkov, M. A. Kabeshov, M. Bella, O.
Kysilka, D. A. Mayshev, K. Pluha´cˇkova´ and P. Kocˇovsky´, Org. Lett.,
2007, 9, 5473–5476; (b) A. Russo and A. Lattanzi, Eur. J. Org. Chem.,
2008, 2767–2773; (c) T. E. Kristensen, K. Vestli, F. K. Hansen and T.
Hansen, Eur. J. Org. Chem., 2009, 5185–5191.
10 N. Halland, P. S. Aburel and K. A. Jørgensen, Angew. Chem., Int. Ed.,
2004, 43, 1272–1277.
11 Recently, a novel oxazolidine has been used as an organocatalyst for
the first time: H. Nakano, K. Osone, M. Takeshita, E. Kwon, C. Seki,
H. Matsuyama, N. Takano and Y. Kohari, Chem. Commun., 2010, 46,
4827–4829.
12 At shorter reaction times (3 days), a decrease in yield (21%) is observed,
although the ee was higher (89%). The use of ethyl benzoylacetate
instead of benzyl benzoylacetate as starting material with 2 as catalyst
provided lower yield (61%) and ee (44%). Comparable results were
previously reported by Jørgensen and coworkers (ref. 10).
13 (a) N. Halland, R. G. Hazell and K. A. Jørgensen, J. Org. Chem.,
2002, 67, 8331–8338; (b) N. Halland, P. S. Aburel and K. A. Jørgensen,
Angew. Chem., Int. Ed., 2003, 42, 661–665; (c) N. Halland, T. Hansen
and K. A. Jørgensen, Angew. Chem., Int. Ed., 2003, 42, 4955–4957;
21 S. M. Allin, S. L. James, M. R. J. Elsegood and W. P. Martin, J. Org.
Chem., 2002, 67, 9464–9467.
22 Gaussian 03, Revision E.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel,
G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr.,
T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchsian, 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, P. Y.
Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G.
Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K.
Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q.
Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P.
M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez and J. A.
Pople, Gaussian, Inc., Wallingford CT, 2004.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 5079–5085 | 5085
©