Asymmetric organocatalytic aldol reaction in water: Mechanistic insights and development of a semi-continuously operating process

Article abstract of DOI:10.1055/s-0033-1338509

A detailed study on the impact of the catalyst loading on conversion and enantioselectivity of the direct aldol reaction of an aldehyde and acetone in aqueous solvent with nonimmobilized and immobilized proline amide-based organocatalysts is described. Based on a polymer-supported catalyst, batch and semi-continuously operating processes, comprising recycling studies in the latter case, were investigated.

Full text of DOI:10.1055/s-0033-1338509

2512▌
FEATURE ARTICLE
feature article
Asymmetric Organocatalytic Aldol Reaction in Water: Mechanistic Insights
and Development of a Semi-Continuously Operating Process
Enantioselective Aldol Reaction Using Proline Amide-Based Catalysts
Giuseppe Rulli,a Kim A. Fredriksen,b Nongnaphat Duangdee,c Tore Bonge-Hansen,b Albrecht Berkessel,*c
Harald Gröger*a,d
a
Department of Chemistry and Pharmacy, University of Erlangen-Nürnberg, Henkestraße 42, 91054 Erlangen, Germany
b
Department of Chemistry, University of Oslo, P. O. Box 1033 Blindern, 0315 Oslo, Norway
c
Department of Chemistry, University of Cologne, Greinstr. 4, 50939 Cologne, Germany
E-mail: berkessel@uni-koeln.de
d
Faculty of Chemistry, Bielefeld University, Universitätsstr. 25, 33615 Bielefeld, Germany
Fax +49(521)1066146; E-mail: harald.groeger@uni-bielefeld.de
Received: 06.04.2013; Accepted after revision: 29.06.2013
Dedicated to Professor Dr. Jürgen Martens on the occasion of his 65th birthday
The advantages of this organocatalytic aldol reaction such
Abstract: A detailed study on the impact of the catalyst loading on
as beneficial catalytic properties, the use of readily acces-
conversion and enantioselectivity of the direct aldol reaction of an
sible substrates (without the need of protecting groups),
aldehyde and acetone in aqueous solvent with nonimmobilized and
immobilized proline amide-based organocatalysts is described.
and availability of an attractive, optimized access to the
Based on a polymer-supported catalyst, batch and semi-continuous- catalyst5 encouraged us to conduct deeper kinetic and
ly operating processes, comprising recycling studies in the latter
case, were investigated.
mechanistic investigations for a detailed characterization
of the reaction course.
Key words: aldol reaction, catalysis, enantioselectivity, supported
Since a range of process features of this aldol reaction al-
catalysis, polymers
ready fulfill the technical requirements, we also became
interested in further process development and the set-up
of a semi-continuously operating process. For the latter is-
The organocatalytic aldol reaction is commonly known as
sue we planned to use the polymer-supported organocata-
the most essential tool in synthetic organic chemistry for
lyst 2, which was developed by the Hansen group
the production of β-hydroxy ketones.1 In the last decade,
recently.6 Furthermore, in such a study the reaction course
an increasing number of catalysts for this reaction have
been developed. However, only a few of them are able to
in the presence of immobilized catalysts 2 should be com-
pared with the one of related nonimmobilized organocat-
alysts 1 (Figure 1). The required immobilized catalysts of
type 2 are accessible by means of a co-polymerization of
catalyze an asymmetric aldol reaction of aldehydes and
acetone in aqueous medium.2 The proline amide-based
catalysts of type 1 developed by the Singh group (Figure
styrene and divinylbenzene with the monomeric proline
1) appear to be among the most suitable types of organo-
amide unit and already showed a promising recycling po-
catalysts for this reaction.3 For example, catalyst 1a in-
tential.6–8
duces high enantioselectivity and is highly active in water,
To start with our study on a detailed characterization of
thus allowing its use at a low catalyst loading of
the asymmetric aldol reaction in aqueous media, the ‘mo-
nomeric’ organocatalyst 1b was first applied. The results
with catalyst 1b also serve as a benchmark for the experi-
ments with the immobilized catalyst 2, which has the
same structural motif of the catalytic moiety. In accor-
dance with previous results collected with the opposite en-
antiomer of catalyst 1a,4 a comparable reaction rate was
obtained when using either 0.5 mol% of organocatalyst 1a
(confirming the previous study) or 0.5 mol% of organo-
catalyst 1b in the aldol reaction of 3-chlorobenzaldehyde
(3; 0.7 M; 0.5 mmol scale) with acetone in a saturated
NaCl solution (Figure 2).
0.5 mol%.3b,4
O
O
O
O
O
R
O
Ph
O
Ph
Ph
Ph
Ph
N
H
N
O
H
NH
OH
NH
OH
1
1a R = i-Bu
1b R = Ph
Figure 1 Proline amide-based organocatalyst 1 and polymer-
supported proline-based organocatalyst 2
Thus, a change of the substituent at the β-amino alcohol
moiety of the organocatalyst 1 (e.g., exchange of an iso-
butyl group by a phenyl group) does not have a significant
impact on the reaction course (Figures 2 and 3). Since the
reactions are kinetically controlled at 0.5 mol% catalyst
loading of 1 and a reaction time of 24 hours, the enantiose-
SYNTHESIS 2013, 45, 2512–2519
Advanced online publication: 12.08.2013
DOI: 10.1055/s-0033-1338509; Art ID: SS-2013-T0274-FA
© Georg Thieme Verlag Stuttgart · New York
0
0
3
9
-
7
8
8
1
1
4
3
7
-
2
1
0
X

FEATURE ARTICLE
Enantioselective Aldol Reaction Using Proline Amide-Based Catalysts
2513
Biographical Sketches█
Giuseppe Rulli was born in Pe-
rugia, Italy, in 1984. He studied
chemistry at the Friedrich Alex-
ander University Erlangen-
Nürnberg and received his di-
ploma degree in 2010. He is
currently conducting his PhD
studies under the supervision of
Prof. Harald Gröger in the field
of asymmetric organocatalysis
and chemoenzymatic syntheses.
Kim Alex Fredriksen was born
in Norway in 1987. He graduat-
ed from University of Oslo in
2011 were he obtained his
Bachelor and Master’s degree
under the supervision of Asso-
ciate Professor Tore Bonge-
Hansen. He is now pursuing a
PhD at the same university
under the supervision of
Associate Professor Mohamed
Amedjkouh.
Nongnaphat Duangdee was
born in Surin, Thailand. She re-
ceived her Bachelor’s degree in
chemistry from Khon Kaen
University in 1996. In 1997, she
joined the Department of Sci-
ence Service (DSS, Thailand)
where she worked as an analyt-
ical chemist until 2009. In par-
allel, she obtained her MSc
(organic chemistry) under the
supervision of Professor Manat
Pohmakotr at Mahidol Univer-
sity, Bangkok in 2002. In 2009,
she started her doctoral studies
under the supervision of Profes-
sor Albrecht Berkessel at the
University of Cologne (Germa-
ny) investigating ‘Organocata-
lytic Asymmetric Aldol Reac-
tions with Ketone Acceptors’
and recently obtained her PhD
in 2013.
Tore Bonge-Hansen was born
in Trondheim, Norway, in
1969. He obtained his PhD from
State University of New York,
Buffalo under the supervision
of Professor Huw M. L. Davies
working on the development of
catalytic asymmetric intermo-
lecular C–H insertion reactions.
He was a postdoctoral fellow in
the laboratory of Professor Karl
Anker Jørgensen in Århus,
Denmark, working on asym-
metric metal-catalyzed reac-
tions and organocatalysis. After
working in the medicinal chem-
istry department at H. Lund-
beck A/S in Copenhagen, he
started his independent research
career as an Associate Professor
at the University of Oslo in
2003. The research in his group
is focused on carbenoid chemis-
try and immobilized organoca-
talysis.
Albrecht Berkessel was born
in Saarlouis, Germany, in 1955.
He obtained his Diplom in 1982
at the University of Saarbrück-
en. For his PhD studies, he
moved to the laboratory of Pro-
fessor Waldemar Adam at the
University of Würzburg. After
receiving his PhD in 1985, he
joined the research group of
Professor Ronald Breslow at
Columbia University, New
York. His habilitation at the
University of Frankfurt/Main
(associated to Professor Ger-
hard Quinkert) was completed
in 1990. In 1992, he became As-
sociate Professor at the Univer-
sity of Heidelberg. Since 1997,
he is a Full Professor of Organic
Chemistry at the University of
Cologne. His research interests
center around various aspects of
catalysis research, such as
mechanism and method devel-
opment in metal-based catalysis
and in particular organocataly-
sis, biomimetic and medicinal
chemistry.
Harald Gröger was born in Er-
langen, Germany, in 1968. He
studied chemistry at the Univer-
sities of Erlangen-Nürnberg and
Oldenburg and received his di-
ploma degree in chemistry from
the University of Oldenburg in
1994. He completed his doctor-
al thesis at the University of
Oldenburg in 1997 under the
supervision of Prof. Dr. Mar-
tens. After staying as a postdoc-
toral fellow at the University of
Tokyo in the group of Prof. Dr.
Shibasaki, he joined the re-
search department Chemische
Forschung of SKW Trostberg
AG in 1998. After the merger
with Degussa-Hüls AG to De-
gussa AG in 2001, he became
project manager in the Project
House Biotechnology of De-
gussa AG. From 2004 to 2006
he worked as a senior project
manager at the research unit
Service Center Biocatalysis of
Degussa AG. From 2006 to
2011 he was W2-professor (as-
sociate professor) for organic
chemistry at the University of
Erlangen-Nürnberg, and since
April 2011 he is a W3-professor
(full professor) for Organic
Chemistry at Bielefeld Univer-
sity. He is the recipient of sever-
al awards, including the
Degussa Innovation Award
2003 (category: new products),
the Degussa Innovation Award
2005 (category: new or im-
proved processes), and the Carl-
Duisberg-Memorial-Prize 2008
of the German Chemical Soci-
ety (GDCh). His main research
areas center on the use of bio-
catalysts in organic synthesis,
comprising, for example, the
development of industrially fea-
sible biotransformations and the
combination of bio- and chemo-
catalysis towards chemoenzy-
matic one-pot processes.
© Georg Thieme Verlag Stuttgart · New York
Synthesis 2013, 45, 2512–2519

2514
G. Rulli et al.
FEATURE ARTICLE
Table 1 Dependency of the Course of the Aldol Reaction on the Cat-
alytic Amount of 1ba
O
O
OH
O
1b
(0.5 to 5.0 mol%)
Cl
Cl
H
sat. aq NaCl
r.t., 24 h
3
(R)-4
0.7 M
9 equiv
Entry 1b (mol%) Overall conv. Product-related conv. ee (%)
(%)b
(%)c
1
2
3
4
0.5
1.0
2.5
5.0
95
93
87
87
91
90
79
73
96
95
80
35
Figure 2 Dependency of the conversion of the aldol reaction on the
reaction time when using organocatalysts 1a and 1b
a Reaction time: 24 h.
b The overall conversion is defined as conversion according to the
consumption of substrate 3.
c The product-related conversion is defined as conversion related to
formation of aldol product (R)-4.
lectivity remains quite constant during the reaction, reach-
ing, for example, 98% ee at the beginning of the reaction
and 96% ee after 24 hours when using 1b (Figure 3).
OH
O
O
OH
O
1a
(0.5 mol%)
Cl
Cl
3
5
sat. aq NaCl
r.t., 24 h
(R)-4
(R)-4
85% ee
89% ee
9 equiv
95% purity
4
3
5
Figure 3 Dependency of the enantioselectivity of the aldol reaction
on the reaction time when using 0.5 mol% of organocatalyst 1a or 1b
(for reaction conditions, see reaction scheme in Figure 2)
Amount in the crude
reaction mixture (%)
91
4
5
OH
O
O
OH
O
1a
(5.2 mol%)
Since we earlier observed that at an elevated catalytic
loading of the opposite enantiomer of 1a (e.g., at 5.0
mol%),4 a thermodynamic control of the process was
reached rapidly, a related reaction with 5.0 mol% of 1b
was conducted. As expected, an increased catalyst loading
of 1b leads to a decrease of enantioselectivity (Table 1).
For example, 5.0 mol% of 1b gives 35% ee after 24 hours
at room temperature.9
Cl
Cl
3
5
sat. aq NaCl
r.t., 24 h
(R)-4
(S)-4
89% ee
9 equiv
5% ee
95% purity
4
3
5
amount in the crude
reaction mixture (%)
76
10
14
As an explanation, we had proposed earlier (when observ-
ing similar results with the opposite enantiomer of catalyst
1a4) the increasing impact of an organocatalytic retro-
aldol reaction at a prolonged reaction time, thus leading to
the thermodynamically favored racemate as the product in
a thermodynamically controlled process.4 To find further
support for this hypothesis we became interested in eval-
uating the course of the ee value of product 4 after mixing
enantiomerically enriched product (R)-4 (89% ee) with
acetone (9 equiv) in the presence of 0.5 mol% and
5.2 mol% of the organocatalyst 1a, respectively (Scheme 1).
O
O
Cl
Cl
H
3
5
Scheme 1 Change of enantiomeric purity of (R)-4 by adding differ-
ent amounts of organocatalyst 1a
when mixed with 0.5 mol% of 1a, a tremendous decrease
of the ee value of product 4 to 5% ee (with preference
even for the opposite S-enantiomer) was found when be-
ing mixed with 5.0 mol% of 1a. The (slight) inversion of
the enantiomeric ratio of 4 in this experiment is postulated
to be related to a different decomposition rate for the two
Any change in enantiomeric composition would indicate
the existence of such a thermodynamic control of the aldol
reaction. In fact, while the enantiomeric excess of product
(R)-4 decreases only slightly to 85% ee after 24 hours
Synthesis 2013, 45, 2512–2519
© Georg Thieme Verlag Stuttgart · New York

FEATURE ARTICLE
Enantioselective Aldol Reaction Using Proline Amide-Based Catalysts
2515
enantiomers of the aldol product 4 under formation of the
undesired condensation by-product 5 in the presence of
the catalyst.
Furthermore, the impact of the reaction temperature on
the aldol reaction of 3-chlorobenzaldehyde (3) with ace-
tone was studied in the presence of 0.5 mol% or 5.0 mol%
organocatalyst 1a (Table 2). Notably, when using a cata-
lytic amount of 0.5 mol% within a reaction time of 17.5
hours, the enantiomeric excess of β-hydroxy ketone (R)-4
turned out to be high with 96–99% ee over a broad tem-
perature range between –20 to 25 °C, underlining the ki-
netic control of these reactions. In strong contrast, the use
of 5.0 mol% catalyst causes a significant decrease of en-
antioselectivity when operating at the same reaction time.
This indicates the impact of a thermodynamic control of
the aldol reaction when using a catalytic amount of 5.0
mol% within a reaction time of 17.5 hours independent of
the applied reaction temperature, and even at –20 °C.
Table 2 Temperature-Dependency of Enantioselectivity with Varia-
tion of the Catalytic Amount of 1a
O
O
OH
O
1a
(0.5 to 5.0 mol%)
Cl
Cl
H
sat. aq NaCl
temp, 17.5 h
Figure 4 Reaction of 4 with deuterated acetone and varying amount
of catalyst 1a and 1H NMR spectra (extracted parts) of substance 4 be-
fore and after the reaction with deuterated acetone in acetone-d6 as a
solvent: spectrum a): reference compound prior to the reaction; spec-
trum b): compound after reaction with 0.5 mol% organocatalyst 1a;
spectrum c): compound after reaction with 5.2 mol% organocatalyst
1a.
3
(R)-4
0.7 M
9 equiv
Entry
1a (mol%) Temp (°C)
Overall conv. (%) ee (%)
1
2
3
4
6
7
8
9
0.5
0.5
0.5
0.5
5.0
5.0
5.0
5.0
–20
0
16
47
87
>95
90
92
90
95
99
97
98
96
89
71
75
54
10
25
–20
0
the formation of a deuterated analogue of aldol product 4.
Exactly this observation was made at a high catalyst load-
ing of 5.2 mol% after 24 hours, whereas a negligible de-
pletion of nondeuterated product 4 was found when
adding 0.5 mol% of organocatalyst 1a (Figure 4). At the
same time, the signal intensity of possibly released ace-
tone increases at higher catalyst amount. These results in-
dicate a significant impact of a retro-aldol reaction of
aldol adduct 4 within 24 hours when operating at a cata-
lytic amount of 5.2 mol%. Nevertheless, with these exper-
iments alone, exchange of α-acidic protons via enamine
formation and subsequent protonation as an alternative
explanation for this effect cannot be excluded (although in
this case at a high percentage partially deuterated products
would have been expected due to statistic reasons, which,
however, were not observed).
10
25
A further option to find support for the proposed retro-
aldol reaction (at higher catalyst loading) as an explana-
tion for the racemization of β-hydroxy ketone 4 consists in
treating a mixture of 4 and catalyst 1a with deuterated
acetone-d6 and monitoring the reaction course via H
NMR spectroscopy (Figure 4).
1
In the case of operating under thermodynamic control,
retro-aldol reaction proceeds, thus forming aldehyde 3 in
situ. Vice versa, in a thermodynamic equilibrium of the al-
dol reaction the aldehyde 3 is then reconverted to 4. Since
in this experiment acetone-d6 is used in excess (9 equiv),
the probability of the aldol reaction of aldehyde 3 is higher
with acetone-d6 compared to acetone, which is formed in
situ. Thus, in the case of a thermodynamic control and the
presence of a retro-aldol reaction, we then should observe
a decrease of nondeuterated aldol adduct 4 (used as a start-
ing material) in the 1H NMR spectrum, corresponding to
With this kinetic and mechanistic picture of the enantiose-
lective aldol reaction with monomeric ‘homogeneous’ or-
ganocatalysts 1 in hand, we became interested in
analogous aldol reactions with related heterogeneous or-
ganocatalysts. A particular focus was on the study of their
impact on enantioselectivity and conversion of aldol prod-
ucts as a function of time and catalyst loading. Heteroge-
neous catalysts offer advantages such as easy separation
from the reaction mixture and simple reuse of the catalyst.
As such heterogeneous immobilized organocatalysts
© Georg Thieme Verlag Stuttgart · New York
Synthesis 2013, 45, 2512–2519

2516
G. Rulli et al.
FEATURE ARTICLE
polymer-supported derivatives of proline amides of type 1
attracted our attention due to their ability to catalyze high-
ly enantioselective aldol reactions in an efficient way.6–8,10a
According to the literature, catalyst loading of most im-
mobilized proline amide organocatalysts does not seem to
have any impact on enantioselectivity within a reaction
time of up to 24–95 hours.6,10b
Table 3 Comparison of Enantioselectivity with Variation of Catalyst 2
O
O
OH
O
2
(0.5 to 50.0 mol%)
Cl
Cl
H
sat. aq NaCl
r.t., 24 h
3
(R)-4
0.7 M
9 equiv
Entry 2 (mol%) Overall conv. Product-related conv. ee (%)
Indeed, when choosing again the aldol reaction of alde-
hyde 3 with acetone in aqueous medium as a model reac-
tion, the catalytic amount of 2 can be increased up to
10.0 mol% (catalyst loading 0.65 mmol/g) without any
loss of enantioselectivity of the resulting product 4 within
24 hours (Figure 5). Accordingly, the reactivity of the het-
erogeneous catalysts of type 2 is much lower compared to
their homogeneous ‘monomeric’ counterparts 1. This is
also supported when comparing the conversion versus re-
action time-curves. For example, the reaction course with
10.0 mol% of the heterogeneous catalyst 2 (Figure 5) is
similar to the one obtained when using 0.5 mol% of the re-
lated nonimmobilized catalyst 1b (see Figure 2). Thus,
within a typical reaction time of 24 hours, the aldol reac-
tions based on the use of the heterogeneous, immobilized
organocatalyst 2 still run under kinetic control, although
an elevated catalytic amount of up to 10.0 mol% was used
and a high ee value of 93% ee was reached after 24 hours
(Figure 5 and Table 3, entry 6).
(%)
(%)
1
2
3
4
5
6
7
0.5
10
10
11
27
30
58
84
64
95
94
93
94
93
93
74
1.0
11
2.5
28
2.5
32
5.0
61
10.0
50.0a
>95
>95
a Due to the high amount of polymer, the reaction takes place in a gel
as a reaction medium.
Besides being used in a batch mode, the immobilized or-
ganocatalyst 2 offers an opportunity to be used for a
(semi-)continuously operating process in a packed-bed re-
actor, which often is a technically preferred option for a
reactor set-up. Although successful recycling of immobi-
lized catalysts by simple filtration and reuse in several cat-
alytic cycles can also be done in a batch mode,7,12
mechanical abrasion of the immobilisate and loss of cata-
lyst are drawbacks here. These can be often circumvented
by using the immobilized catalyst in a packed-bed reac-
tor.13 In the following, our results based on the use of a
semi-continuous flow reaction is described. In the set-up
of this experiment the polymer-supported organocatalyst
2 represents the stationary phase and was inserted into a
cartridge, which is connected to an adjustable pumping
device (Figure 6). Since pumping the aqueous mixture of
acetone and substrate 3 was considered to be disadvanta-
geous since it represents a biphasic system, at first the cat-
alyst was treated with water to ensure an aqueous
environment of the heterogeneous catalyst 2. Then, the re-
actants 3-chlorobenzaldehyde (3) and acetone were
pumped with a constant flow of 0.5 mL/min through the
packed-bed reactor and recollected in the same vessel that
contains the starting materials. After 18 hours, the reac-
tion was stopped (by interrupting the pumping) and sub-
sequently the reaction mixture was analyzed, showing a
conversion of 93% [with respect to formation of aldol
product (R)-4] and an enantioselectivity of 90% ee (Table
4, entry 1). Notably, in the absence of pretreatment of cat-
alyst 2 with water the conversion was considerably low-
er.14 This underlines the importance of water as an
essential component of the reaction medium for an effi-
cient organocatalytic aldol reaction of aldehyde 3 with
acetone.
Figure 5 Dependency of the conversion and enantioselectivity of the
aldol reaction on the reaction time when using 10.0 mol% of organo-
catalyst 2
Next we conducted a detailed study of the impact of cata-
lyst loading of the heterogeneous organocatalyst 2 on the
enantiomeric excess of the resulting aldol adduct (R)-4,
varying the catalyst loading between 0.5 mol% and 50.0
mol% (Table 3). A decrease of the enantioselectivity was
only observed when using a very large amount of poly-
mer-supported catalyst 2. Accordingly, the use of
50.0 mol% of 2 causes a drop of the enantiomeric excess
of (R)-4 to 74% ee (Table 3, entry 7).11
Synthesis 2013, 45, 2512–2519
© Georg Thieme Verlag Stuttgart · New York

FEATURE ARTICLE
Enantioselective Aldol Reaction Using Proline Amide-Based Catalysts
2517
Table 5 Recycling of Polymer-Supported Organocatalyst 2 in a
Semi-Continuous Flow Reactor
2
O
O
OH
O
packed-bed reactor
(pretreatment with H2O)
Cl
Cl
H
(used as
solvent)
r.t., 24 h
5 reaction cycles
3
(R)-4
Entry
Cycle
Product-related conv. (%)
ee (%)
1
2
3
4
5
1
94
92
92
92
88
88
2
93
Figure 6 Sketch of the semi-continuous-flow reactor for the aldol re-
action (shown in Table 4)
3
93
4a
5a
93
Table 4 Variation of Absolute Amount of Aldehyde 3 for the Aldol
Reaction of Polymer-Supported Organocatalyst 2 in a Semi-Continu-
ous Flow Reactor
>95
a A different pumping device was used for this reaction (see experi-
mental section).
2
O
O
OH
O
packed-bed reactor
(pretreatment with H2O)
Cl
Cl
H
(used as
solvent)
r.t., time (h)
amount of 0.5 mol%. However, the use of the polymer-
supported organocatalyst 2 leads to a kinetically con-
trolled reaction within a broad range of the catalytic
amount of 0.5 to 10.0 mol%. This is due to a lower reac-
tivity of the heterogeneous catalyst. Furthermore, based
on the use of the immobilized catalyst 2 a proof of concept
was demonstrated for a semi-continuously operating or-
ganocatalytic aldol reaction with a packed-bed reactor
bearing the organocatalyst 2 as a stationary phase. When
employed in such a semi-continuous flow reactor the im-
mobilized catalyst 2 also showed a promising recycling
potential, giving high conversion and enantioselectivities
for the aldol product (R)-4 within five reaction cycles.
3
(R)-4
Entry 3 (mmol) Time (h) Product-related conv. (%) ee (%)
1
0.5
8.8
18
66
93
95
90
89
2a
a Polymer-supported organocatalyst 2 was used two times (in total
36 h of previous semi-continuous use) before starting this reaction.
Furthermore, the organocatalyst 2 exhibits a promising re-
cycling potential when being employed in the packed-bed
reactor. High conversions (up to >95%) and high enanti-
oselectivities (88–92% ee) were observed over five cy-
cles, each lasting at least 24 hours (Table 5). It should be
added that pretreatment of the stationary phase (consisting
of immobilized catalyst 2) with water was done prior to
each reaction cycle.
All reagents were purchased from commercial sources and used
without further purification, unless otherwise indicated. Acetone
was distilled before use. 3-Chlorobenzaldehyde (3) was distilled be-
fore use (if its purity was less than 97%). Organocatalysts 1a, 1b,
and 2 were synthesized in analogy to literature-known proce-
dures.3a,4–7 1H NMR spectra were recorded on a Bruker Avance 400
and 300 spectrometer and the chemical shift values refer to CDCl3
[δ (1H), 7.24 ppm] and acetone-d6 [δ (1H), 2.05 ppm]. HPLC was
carried out with a Jasco LCNet II/ADC using a UV-1575 UV/Vis
Detector and Daicel Chiralpak® column AD-H. Low-temperature
reactions were carried out with a JULABO typ/model FT902
Kryostat. The packed-bed reactor consists of a Chiralpak® column
cartridge (Ø = 4.6 mm, length = 250 mm), with organocatalysts 2
included therein (particle size: ca. 200 μm) connected to A Shimad-
zu LC-10AT or alternatively to a Jasco LCNet II pumping device
(see Table 5, entries 4 and 5). Conversions were determined from
the ratio of the integral of product signals in the 1H NMR spectra of
the crude products in relation to the corresponding sum of integrals
In addition, reutilization of 2 (after 36 h of previous semi-
continuous use, as shown in Table 4) in combination with
an 18-fold amount of 3-chlorobenzaldehyde (3) and a pro-
longed reaction time of 66 hours leads to the formation of
the desired β-hydroxy ketone (R)-4 with 95% conversion
and 89% ee (Table 4, entry 2).
In summary, a detailed kinetic and mechanistic study for
the characterization of the reaction course of the enantio-
selective direct aldol reaction catalyzed by ‘monomeric’ of substrate (aldehyde), product and clearly identifiable side-prod-
ucts (the synthesis of compound 5 as a reference is given herein, the
synthesis of a second by-product is given in the Supporting Infor-
proline amide-based catalysts 1 was conducted. The re-
sults indicate that an elevated catalytic amount of 5.0
mation). Enantioselectivities were determined by chiral HPLC
mol% in combination with a reaction time of 24 hours led
to a significant impact of the retro-aldol reaction and a
thermodynamic control of the reaction. In contrast, excel-
lent enantioselectivities (at high conversions) were ob-
tained within this reaction time at a decreased catalytic
(Chiralpak® column AD-H, hexane–i-PrOH, 95:5, flow 1.0
mL/min, 220 nm), after complete removal of the solvent, directly
from the crude product, without further purification. The analytical
data obtained for the synthesized compound 4 were in accordance
with the corresponding literature-known data.3b,4
© Georg Thieme Verlag Stuttgart · New York
Synthesis 2013, 45, 2512–2519

2518
G. Rulli et al.
FEATURE ARTICLE
Organocatalytic Aldol Reaction (Figures 2 and 3); General
Procedure
The resulting crude product was then analyzed with respect to con-
version as well as enantiomeric excess of (R)-4.
A mixture of 3-chlorobenzaldehyde (3; 59 μL, 0.5 mmol), organo-
catalyst 1a (0.9 mg, 0.0025 mmol, corresponding to a catalytic
amount of 0.5 mol%) or organocatalyst 1b (1.0 mg, 0.0025, corre-
sponding to a catalytic amount of 0.5 mol%), and acetone (329 μL,
4.5 mmol, 9 equiv) in a sat. aq solution of NaCl (0.33 mL, corre-
sponding to about 50% v/v) was stirred at r.t. (for reaction times, see
Figures 2 and 3). The reaction mixture was extracted with CH2Cl2
(3 × 2 mL) and after the organic layers were combined, the solvent
was removed by evaporation (600 mbar, 40 °C). The resulting crude
product was then analyzed with respect to conversion as well as en-
antiomeric excess of (R)-4.
Organocatalytic Retro-Aldol Reaction (Figure 4); General
Procedure
A mixture of β-hydroxy ketone 4 (100 μL, 103.9 mg, 95% purity,
corresponding to 0.48 mmol), organocatalyst 1a (0.9–9.1 mg,
0.0025–0.025 mmol, corresponding to a catalytic amount of 0.5–
5.2 mol%), and acetone-d6 (329 μL, 4.4 mmol, 9 equiv) in a sat. aq
solution of NaCl (0.33 mL) was stirred at room temperature for 24
h. The organic layer was separated and analyzed via 1H NMR spec-
troscopy.
Organocatalytic Aldol Reaction (Figure 5); General Procedure
A mixture of 3-chlorobenzaldehyde (3, 59 μL, 0.5 mmol), organo-
catalyst 2 (76.9 mg, catalyst loading 0.65 mmol/g, corresponding to
a catalytic amount of 10.0 mol%), and acetone (329 μL, 4.5 mmol,
9 equiv) in a sat. aq solution of NaCl (0.33 mL, corresponding to
about 50% v/v) was stirred at r.t. (for reaction times, see Figure 5).
The reaction mixture was diluted with CH2Cl2 (5 mL) and the phas-
es were separated over a phase-separating filter. The solvent of the
organic filtrate was removed by evaporation (600 mbar, 40 °C). The
resulting crude product was then analyzed with respect to conver-
sion as well as enantiomeric excess of (R)-4.
Organocatalytic Aldol Reaction (Table 1); General Procedure
A mixture of 3-chlorobenzaldehyde (3; 59 μL, 0.5 mmol), organo-
catalyst 1b (1.0–9.6 mg, 0.0025–0.025 mmol, corresponding to a
catalytic amount of 0.5–5.0 mol%), and acetone (329 μL, 4.5 mmol,
9 equiv) in a sat. aq solution of NaCl (0.33 mL, corresponding to
about 50% v/v) was stirred at r.t. for 24 h. The reaction mixture was
extracted with CH2Cl2 (3 × 2 mL) and after the organic layers were
combined, the solvent was removed by evaporation (600 mbar,
40 °C). The resulting crude product was then analyzed with respect
to conversion as well as enantiomeric excess of (R)-4.
Organocatalytic Aldol Reaction (Table 3); General Procedure
A mixture of 3-chlorobenzaldehyde (3; 59 μL, 0.5 mmol), organo-
catalyst 2 (3.9–384.6 mg, catalyst loading 0.65 mmol/g, corre-
sponding to a catalytic amount of 0.5–50.0 mol%), and acetone
(329 μL, 4.5 mmol, 9 equiv) in a sat. aq solution of NaCl (0.33 mL,
corresponding to about 50% v/v) was stirred at r.t. for 24 h. The re-
action mixture was diluted with CH2Cl2 (5 mL for entries 1–5, 20
mL for entries 6, 7) and the phases were separated over a phase-sep-
arating filter. The solvent of the organic filtrate was removed by
evaporation (600 mbar, 40 °C). The resulting crude product was
then analyzed with respect to conversion as well as enantiomeric
excess of (R)-4.
Detection of Organocatalytic Retro-Aldol Reaction (Scheme 1);
General Procedure
A mixture of β-hydroxy ketone (R)-4 (100 μL, 103.9 mg, 95% pu-
rity, corresponding to 0.48 mmol, 89% ee), organocatalyst 1a (0.9–
9.1 mg, 0.0025–0.025 mmol, corresponding to a catalytic amount of
0.5–5.2 mol%), and acetone (329 μL, 4.5 mmol, 9 equiv) in a sat. aq
solution of NaCl (0.33 mL) was stirred at r.t. for 24 h. The reaction
mixture was diluted with CH2Cl2 (5 mL) and the phases were sepa-
rated over a phase-separating filter. After removing most of the sol-
vent of the organic filtrate by evaporation (600 mbar, 40 °C), the
amount of aldol product (R)-4 or (S)-4 was determined via 1H NMR
from the ratio of the integral of its signals in relation to the sum of
integrals of aldehyde and detectable side-product 5.
Organocatalytic Aldol Reaction (Tables 4, 5); General
Procedure
(E)-4-(3-Chlorophenyl)but-3-en-2-one (5, Scheme 1)
Distilled H2O was filled into a 250 mL glass bottle and pumped with
a flow of 0.5 mL/min through the packed-bed reactor filled with of
organocatalyst 2 [1.32 g (Table 4) or 1.66 g (Table 5), catalyst load-
ing 0.65 mmol/g]. Subsequently, the pump was stopped and the bot-
tle was replaced by a 250 ml glass bottle filled with a mixture of
acetone (25 mL) and 3-chlorobenzaldehyde (3; 59 μL–1.0 mL, 0.5–
8.8 mmol). Under constant stirring, this solution was also pumped
with a flow of 0.5 mL/min into the packed bed reactor and recollect-
ed into the bottle (for reaction times, see Tables 4, 5). Afterwards,
the pumping was stopped and the solution was diluted with CH2Cl2
(25 mL) (Table 4) or extracted (Table 5) with CH2Cl2 (3 × 20 mL)
and decanted over a phase-separation filter. The solvent of the or-
ganic filtrate was removed by evaporation (500 mbar, 40 °C). After
rinsing the reactor with acetone, the system was reused for the next
cycle. The resulting crude product was then analyzed with respect
to conversion as well as enantiomeric excess of (R)-4.
The synthesis of this reference compound 5 was carried out in anal-
ogy to a literature-known protocol;15 yield: 22%; colorless solid; mp
36 °C; Rf = 0.53 [petroleum ether (bp 40–65 °C)–EtOAc, 5:1].
HPLC (Chiralpak AD-H-column; hexanes–i-PrOH, 95:5; 1.0
mL/min): tR = 6.7 min.
IR (film): 3075, 3024, 1664, 16407, 1625, 1391, 1285, 1258, 1089,
980, 783, 686 cm–1.
1H NMR (300 MHz, CDCl3): δ = 7.50–7.30 (m, 5 H), 6.70–6.65 (d,
J = 16.3 Hz, 1 H), 2.36 (s, 3 H).
13C NMR (75 MHz, CDCl3): δ = 198.0, 141.6, 136.2, 134.9, 130.3,
130.2, 128.1, 127.9, 126.4, 27.7.
MS (EI): m/z = 180 ([M+], 100%).
Anal. Calcd for C10H9ClO: C, 66.49; H, 5.02. Found: C, 66.69; H,
5.11.
Organocatalytic Aldol Reaction (Table 2); General Procedure
A mixture of 3-chlorobenzaldehyde (3; 59 μL, 0.5 mmol), organo-
catalyst 1a (0.9–9.1 mg, 0.0025–0.025 mmol, corresponding to a
catalytic amount of 0.5–5.0 mol%), and acetone (329 μL, 4.5 mmol,
9 equiv) in a sat. aq solution of NaCl (0.33 mL, corresponding to
about 50% v/v) was stirred for 17.5 h (for reaction temp, see Table
2). The reaction mixture was diluted with CH2Cl2 (5 mL) and the
phases were separated over a phase-separating filter. The solvent of
the organic filtrate was removed by evaporation (600 mbar, 40 °C).
Acknowledgment
We thank the Deutsche Forschungsgemeinschaft (DFG) for ge-
nerous support within the priority programme SPP1179 ‘Organoka-
talyse’ (BE 998/11-1, GR 3461/2-1). Membership in the European
Cooperation in Science and Technology (COST) Action no.
CM0905 Organocatalysis (ORCA) is gratefully acknowledged.
Synthesis 2013, 45, 2512–2519
© Georg Thieme Verlag Stuttgart · New York

FEATURE ARTICLE
Enantioselective Aldol Reaction Using Proline Amide-Based Catalysts
2519
(9) The overall conversions decrease slightly by applying higher
catalyst loadings (e.g., 1.0 mol%: 93%; 5.0 mol%: 87%).
Although no detailed explanation is as yet available, this
effect might be related (at least to some extent) to an
increased formation of an inactive catalyst–substrate adduct
in the case of reactions conducted at higher catalyst loading
(for the formation of related adducts, see ref. 3d). The
differences in the product-related conversions are primarily
caused by the higher impact of the side-product formation
(via dehydration of aldol product 4) in the case of the
reactions conducted at higher catalyst loading.
(10) (a) Gruttadauria, M.; Giacalone, F.; Noto, R. Chem. Soc.
Rev. 2008, 37, 1666. (b) Gruttadauria, M.; Giacalone, F.;
Marculescu, A. M.; Noto, R. Adv. Synth. Catal. 2008, 350,
1397.
(11) Because of the large amount of catalyst, the solvent is
completely absorbed by the polymer, leading to the
formation of a gel.
Supporting Information for this article is available online at
http://www.thieme-connect.com/ejournals/toc/synthesis.SnuIomprifgtooatrinSugpIoi
n
nfomartit
References
(1) (a) Modern Methods in Stereoselective Aldol Reactions;
Mahrwald, R., Ed.; Wiley-VCH: Weinheim, 2013.
(b) Berkessel, A.; Gröger, H. Asymmetric Organocatalysis;
Wiley-VCH: Weinheim, 2005. (c) Melchiorre, P.; Marigo,
M.; Carlone, A.; Bartoli, G. Angew. Chem. Int. Ed. 2008, 47,
6138; Angew. Chem. 2008, 120, 6232.
(2) (a) Tang, Z.; Yang, Z.-H.; Chen, X.-H.; Cun, L.-F.; Mi, A.-
Q.; Jiang, Y.-Z.; Gong, L.-Z. J. Am. Chem. Soc. 2005, 127,
9285. (b) Raj, M.; Singh, V. K. Chem. Commun. 2009, 6687.
(3) (a) Raj, M.; Maya, V.; Ginotra, S. K.; Singh, V. K. Org. Lett.
2006, 8, 4097. (b) Maya, V.; Raj, M.; Singh, V. K. Org. Lett.
2007, 9, 2593. (c) Raj, M.; Maya, V.; Singh, V. K. J. Org.
Chem. 2009, 74, 4289. (d) Duangdee, N.; Harnying, W.;
Rulli, G.; Neudörfl, J.-M.; Gröger, H.; Berkessel, A. J. Am.
Chem. Soc. 2012, 134, 11196. (e) Baer, K.; Kraußer, M.;
Burda, E.; Hummel, W.; Berkessel, A.; Gröger, H. Angew.
Chem. Int. Ed. 2009, 48, 9355; Angew. Chem. 2009, 121,
9519.
(4) Rulli, G.; Duangdee, N.; Baer, K.; Hummel, W.; Berkessel,
A.; Gröger, H. Angew. Chem. Int. Ed. 2011, 50, 7944;
Angew. Chem. 2011, 123, 8092.
(5) Berkessel, A.; Harnying, W.; Duangdee, N.; Neudörfl, J.-
M.; Gröger, H. Org. Process Res. Dev. 2011, 16, 123.
(6) Kristensen, T. E.; Vestli, K.; Fredriksen, K. A.; Hansen, F.
K.; Hansen, T. Org. Lett. 2009, 11, 2968.
(7) Kristensen, T. E.; Vestli, K.; Jakobsen, M. G.; Hansen, F. K.;
Hansen, T. J. Org. Chem. 2010, 75, 1620.
(8) Kristensen, T. E.; Hansen, T. Eur. J. Org. Chem. 2010, 3179.
(12) An example for the recycling of 2 in a batch reactor is given
in the Supporting Information.
(13) For recent applications of immobilized organocatalysts in
continuous flow systems, see: (a) Ötvös, S. B.; Mándity, I.
M.; Fülöp, F. ChemSusChem 2012, 5, 266. (b) Ayats, C.;
Henseler, A. H.; Pericàs, M. A. ChemSusChem 2012, 5, 320.
(c) Arakawa, Y.; Wennemers, H. ChemSusChem 2013, 6,
242.
(14) (a) Rulli, G.; Gröger, H., unpublished results. (b) For further
information about the role of water as an essential
component for an efficient organocatalytic aldol reaction,
see ref. 2b and: Gruttadauria, M.; Giacalone, F.; Noto, R.
Adv. Synth. Catal. 2009, 351, 33.
(15) Li, X.; Li, L.; Tang, Y.; Zhong, L.; Cun, L.; Zhu, J.; Liao, J.;
Deng, J. J. Org. Chem. 2010, 75, 2981.
© Georg Thieme Verlag Stuttgart · New York
Synthesis 2013, 45, 2512–2519