Efficient synthesis of supported proline catalysts for asymmetric aldol reactions

Article abstract of DOI:10.1039/c4cy00970c

Proline has been grafted onto silica supports in a single step by reacting trans-4-hydroxy-L-proline with chloropropyl tethers, without the use of protecting groups for the proline amine and carboxylic acid functional groups. The resulting catalysts have

Full text of DOI:10.1039/c4cy00970c

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Efficient synthesis of supported proline catalysts
for asymmetric aldol reactions†
Cite this: DOI: 10.1039/c4cy00970c
A. A. Elmekawy, J. B. Sweeney and D. R. Brown*
Proline has been grafted onto silica supports in a single step by reacting trans-4-hydroxy-L-proline with
chloropropyl tethers, without the use of protecting groups for the proline amine and carboxylic acid
functional groups. The resulting catalysts have been characterised to show that grafting is through reaction
with the 4-hydroxy group. The catalysts have been tested in an asymmetric aldol reaction, and shown to
be both more active and more enantioselective than equivalent catalysts prepared using a protection/
deprotection route for the proline grafting step.
Received 25th July 2014,
Accepted 11th September 2014
DOI: 10.1039/c4cy00970c
www.rsc.org/catalysis
examples of proline-bearing silica materials in the literature
Introduction
1
7,18
19
17
20
(
including MCM-41,
SBA-15, silica gel, and zeolites ).
Enantiomerically-pure substances are high added-value
chemicals and have been employed in a diverse range of
The first report of the use of proline–grafted silica to catalyse
asymmetric aldol synthesis appeared in 2003.
1
7
1
,2
research fields, including flavour and aroma chemicals,
The majority of the reported methods to prepare silica-
supported proline catalysts describe relatively complex, multi-
step syntheses of the solid catalysts, incorporating protection/
deprotection steps for the two functional groups on the
proline. Typically, 4-hydroxy-L-proline is used as the starting
material (Fig. 1b). Protection of the secondary amine is
achieved with t-butyloxycarbonyl (Boc), and the carboxylic
acid is protected by esterification (Fig. 1c). A recent publi-
cation reporting the use of unprotected proline in the prep-
3
,4
5
nonlinear optical devices, agricultural chemicals and the
6
manufacture of pharmaceuticals. Many modern drug sub-
stances are chiral and contemporary regulatory requirements
demand high levels of enantiopurity in clinical trials. There
has, therefore, been a long-standing interest in synthetic
methods which deliver non-racemic products, and there has
been a recent rapid development in catalytic enantioselective
synthetic methodologies.
2
1
The majority of methods reported to date have involved
aration of a polymer-supported proline organocatalyst
7
,8
homogeneous systems. There has been recent interest in the
use of proline-derived chiral catalysts, and the original seminal
has stimulated the work described in this paper, in which
a silica gel surface is functionalised with proline, bypassing
the protection/deprotection steps. The object has been to
support proline on silica in a single step using trans-4-
hydroxy-L-proline, by reacting a chloro alkyl tether on the
silica with the hydroxy group on the proline. We aim to
show how this can be achieved without compromising the
catalytic properties (and associated enantioselectivity) of the
two functional groups (Scheme 1).
9
,10
observations of an asymmetric annelation processes
been greatly expanded in scope more recently.
(
have
Proline
Fig. 1a) is bifunctional, with carboxylic acid and amine
1
1–13
functionalities. These two functional groups can, in princi-
ple, act in concert in cooperative acid/base mechanisms, or
can act sequentially, catalysing tandem reaction steps.
Significant advantages in terms of catalyst separation and
recovery can be delivered if proline-derived organocatalysts
are immobilized on solid supports, and a number of catalysts
of this type have been reported, mainly using polymer-
The effect of the physical nature of the support is also
investigated, by basing catalysts on an ordered mesoporous
2
−1
silica (MCM-41, 1000 m g ) and on two commercial amor-
phous silica gels.
1
4–16
mounted systems.
However, silica is a more attractive
support material because of its thermal and mechanical sta-
bility as well as its chemical inertness, and there are several
Department of Chemical Sciences, University of Huddersfield, Huddersfield HD1
3
DH, UK. E-mail: d.r.brown@hud.ac.uk; Fax: +44 (0)1484 472182;
Tel: +44 (0)1484 47339
Electronic supplementary information (ESI) available: HPLC chromatograms,
†
NMR spectra, kinetic data for catalytic measurements. See DOI: 10.1039/
c4cy00970c
Fig. 1 a) L-Proline, b) trans-4-hydroxy-L-proline, c) N-Boc-trans-4-
hydroxy-L-proline methyl ester.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
silica surface. Nitrogen and chlorine levels in the catalysts
were determined by elemental analysis (MEDAC Ltd.). Infrared
spectra of catalysts were recorded using attenuated total reflec-
−
1
tance on a Nicolet 380 FTIR spectrometer over 400–4000 cm .
Catalyst synthesis
Chloropropyl-functionalized silica gels (SiO
2
-Cl)
Scheme 1 Reaction between IJ3-chloropropyl) trimethoxysilane bound
to a silica surface and trans-4-hydroxy-L-proline.
These catalysts, as shown in Scheme 1, were synthesized
2
2
using a modification of a published procedure. The two
amorphous silica gels were used. Silica gel (1.0 g) was added
to toluene (40 mL). The mixture was heated to 110 °C and
stirred for 1 h under nitrogen. IJ3-chloropropyl)trimethoxysilane
(
2.0 mmol, 0.4 mL) was added under stirring at the same
temperature under nitrogen. After 12 h the solid was filtered,
washed with toluene and then dried in an oven at 120 °C in
air for 12 h. The material prepared from silica gel with low
Scheme 2 Cross-aldol reaction between 4-nitrobenzaldehyde (1) and
acetone to give the aldol product (2).
2
surface area was labelled SiO -Cl (L) and the material with
high surface area SiO -Cl (H).
2
The catalysts are tested in a direct asymmetric mixed
aldol reaction, between 4-nitrobenzaldehyde (1) and acetone
Proline-functionalized silica gel (SiO -Pr) by the direct route
2
(
Scheme 2). In this reaction, acetone reacts with L-proline to
2
SiO -Cl (1.0 g) was suspended in xylene (20 mL) at 140 °C
generate a chiral enamine, which then adds to the aldehyde
CO bond, leading eventually to the (S)-configured aldol.
The reaction proceeds via an intermediate chiral iminium,
which liberates the (S)-aldol product, regenerating the proline
under nitrogen and stirred for 1 h. Trans-4-Hydroxy-L-proline
(2 mmol, 0.26 g) was added under stirring at the same tem-
perature under nitrogen. After 12 h, the solid catalyst was
filtered, washed with water and ethanol, and dried in an oven
1
1
catalyst.
2
at 120 °C in air for 12 h to give SiO -Pr. The two catalysts
using the low and the high surface area silica gel supports
were labelled SiO -Pr (L) and SiO -Pr (H).
Experimental
2
2
Materials and instrumentation
Proline-functionalized silica gel by the amine
protecting/deprotecting route
Tetraethyl orthosilicate (TEOS, 98%), IJ3-chloropropyl)trimethoxy-
silane, trans-4-hydroxy-L-proline, N-Boc-trans-4-hydroxy-L-proline,
trifluoroacetic acid (99%), N-Boc-trans-4-hydroxyproline
methyl ester, cetyltrimethylammonium bromide (CTAB) as a
structure directing agent for synthesis of MCM-41, and two
SiO -Cl (L) (1.0 g) was suspended in xylene (20 mL) at 140 °C
2
under nitrogen and stirred for 1 h. Boc-trans-4-hydroxy-L-proline
(
2 mmol, 0.46 g) was added under stirring at the same tem-
perature under nitrogen. After 12 h the reaction mixture was
filtered. The solid was washed with water and ethanol, and
commercial mesoporous silica gels (SiO (H), with surface
2
2
−1
area 516 m g and average pore diameter 4.6 nm, and SiO
2
2
−1
dried as before to give SiO
2
-Pr-Boc. To remove the Boc group,
(
1
L), with surface area 312 m g and average pore diameter
1 nm) were all purchased from Aldrich and used as received.
SiO -Pr-Boc (0.5 g) was suspended in CH Cl (5.0 mL) for
2
2
2
1
0 min, then trifluoroacetic acid (99%) (10 mL) was added to
All reagents for the catalytic tests were also purchased from
Aldrich and used as received.
Powder X-ray diffraction (XRD) patterns were measured on
a Rigaku Miniflex diffractometer over a 2θ range of 1.5°–10°
2
3
the solution. The mixture was left for 12 h at room temper-
ature. The solid was filtered, washed with water and then eth-
2
anol, and dried at 120 °C to give SiO -Pr (Boc)deprotected.
−
1
with a step size of 0.05° and a counting time of 30 s degree .
Proline-functionalized silica gel by the amine and carboxylic
acid protecting/deprotecting route
The N adsorption–desorption isotherms were measured at
2
7
7 K on a Micromeritics ASAP-2020, after evacuation at 423 K
for 5 h. Surface areas were calculated by the BET method
from the adsorption branch and pore size distributions were
determined from the desorption branch using the BJH
method. Thermogravimetric analysis (TGA) was carried out
using a Perkin-Elmer thermobalance. The materials were
heated to 150 °C and held at this temperature for 2 h to
ensure evaporation of all surface water. The temperature was
2
SiO -Cl (L) (1.0 g) was suspended in xylene (20 mL) at 140 °C
under nitrogen and stirred for 1 h. Then N-Boc-trans-4-
hydroxyproline methyl ester (2 mmol, 0.5 g) was added under
stirring at the same temperature under nitrogen. After 12 h,
the proline catalyst was filtered, washed with water and etha-
nol, and then dried at 120 °C to give SiO -Pr-Boc-Me. For
2
2
deprotection, SiO -Pr-Boc-Me (0.5 g) was suspended in 30%
−
1
24
increased to 500 °C at 10 °C min under nitrogen to
estimate the concentration of organic groups attached to the
HBr in acetic acid (10 mL). The mixture was left for 12 h at
room temperature then the solid was filtered, washed with
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
water and ethanol and then dried in an oven at 120 °C in air
removed by filtration and the filtrate solution was tested for
catalytic activity in the aldol reaction.
for 12 h to give SiO -Pr (Boc, Me)deprotected.
2
Proline-functionalized MCM-41 silica by the direct route
Results and discussion
(MCM-41-Pr)
The powder XRD patterns for MCM-41, MCM-41-Cl and
MCM-41-Pr are presented in Fig. 2. They show characteristic
reflections that can be indexed on a hexagonal unit cell to
Mesoporous silica (MCM-41) was synthesized according to a
literature procedure using cetyltrimethylammonium bromide
2
5
as template. For a typical synthesis, 65% NH
4
OH (69 mL)
26
(
100), (110) and (200) by analogy with literature reports.
was mixed with water (525 mL). CTAB (0.125 g) was added
with stirring at 80 °C. When the solution became homoge-
nous, TEOS (1.0 g) was added. After 2 h, the resulting product
was filtered, washed with water, dried at ambient tempera-
ture, and calcined in air at 540 °C for 4 h to obtain MCM-41.
Covalent grafting of the proline derivative onto MCM-41 was
carried out as outlined for the amorphous silica by reacting
IJ3-chloropropyl) trimethoxysilane in toluene at 110 °C for
This data confirms the preservation of the ordered structure
throughout the grafting and proline modification procedures.
Nitrogen adsorption–desorption isotherms for the MCM-41
support, and for MCM-41-Cl and MCM-41-Pr are shown in
Fig. 3, illustrating the effect of functionalisation on surface
properties. All three materials display type IV isotherms, indi-
27
cating the presence of mesopores. With the introduction of
organic functional groups, the surface areas and pore vol-
umes progressively decrease, implying successful grafting of
these groups in the pores of the support. Table 1 summarises
the data derived from the nitrogen adsorption–desorption
experiments for the prepared materials.
1
2 h, followed by direct reaction with trans-4-hydroxy-L-proline
in xylene at 140 °C for 12 h. The solid was filtered, washed
and dried as before to give MCM-41-Pr.
General procedure for asymmetric aldol reaction
The thermogravimetric curve for SiO -Cl (H) is given in
2
Fig. 4. This example is shown to illustrate how TGA data has
been used to characterise the extent of functionalisation of
The catalyst (0.05 g), 4-nitrobenzaldehyde (151 mg, 1 mmol),
acetone (10 mL) and benzyl alcohol (0.10 mL) as internal
standard were placed in a 50 mL round-bottomed flask and
stirred at 50 °C for 3 h under nitrogen. When testing trans-4-
hydroxy-L-proline in homogeneous solution, 0.19 mmol were
used. This corresponds to three to five times the amount of
proline that was present on the 0.05 g of supported catalyst.
Samples (100 μL) were removed at regular intervals, and the
%
conversion of nitrobenzaldehyde and aldol product yield
were estimated by analysis on a Shimadzu HPLC instrument
using a C18 column (water/xylene = 70/30) with 254 nm UV
detection. The identity of reaction products were confirmed
with NMR (see ESI†). The enantiomeric excess (%) of the
aldol product was determined by HPLC analysis using a Lux
Cellulose-3 column (0.1% formic acid in hexane/isopropanol =
9
0/10) with 254 nm UV detection after re-dissolving the
Fig. 2 Powder X-ray diffraction patterns of MCM-41, MCM-41-Cl and
product in isopropanol. All kinetic experiments were run at
least twice and usually three or more times, and 95% confi-
dence limits for conversions, yields and enantiomeric
excesses were estimated from the replicate data.
MCM-41-Pr.
Catalytic tests were also performed using benzaldehyde,
4
-chlorobenzaldehyde and 4-isopropylbenzaldehyde in
place of 4-nitrobenzaldehyde. Experiments with 4-nitro-
benzaldehyde were also performed using 2.0 mL acetone and
8
.0 mL dimethylsulfoxide (DMSO) in place of 10.0 mL ace-
tone, at the lower reaction temperature of 25 °C, for 20 h.
To assay the retained catalytic activity of the silica-
supported proline organocatalysts, two methods were used.
In the first, the catalyst was separated from the reaction mix-
ture after reaction, washed with acetone, and dried. It was re-
introduced to the reaction vessel. Fresh starting materials
were added and the reaction was run again. In the second
method, the leaching of proline was tested by stirring the
catalyst in acetone at 50 °C for 12 h. The catalyst was then
Fig. 3
2
N adsorption–desorption isotherms at 77 K for MCM-41,
MCM-41-Cl and MCM-41-Pr.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Table 1 Textural properties of catalysts and catalyst precursors
the unreacted chloropropyl tether, a weighted average of the
molecular weights of the two functional groups on each cata-
lyst is determined, and this average molecular weight is then
used to convert the measured weight loss (by TGA) from each
sample over 200–500 °C to an effective molar concentration
of organic groups on the catalyst surface.
These values appear as the “organic content” in Table 2.
These values can be compared with the sum of the molar
concentrations of N and Cl from elemental analysis. On this
basis, it can be seen that the overall concentrations of surface
functional groups from elemental analysis (N plus Cl concen-
trations) agree very well with the overall concentrations based
on TGA.
Surface2
Pore
volume/cm g
Average pore
diameter/nm
−1
3
−1
Sample
area/m g
SiO
SiO
SiO
SiO
SiO
SiO
SiO
SiO
2
2
2
2
2
2
2
2
(L)
-Cl (L)
-Pr (L)
-Pr-Boc (L)
-Pr-Boc,Me (L)
(H)
-Cl (H)
-Pr (H)
313
278
265
260
257
517
361
357
1100
998
959
1.1
0.9
0.8
0.8
0.8
0.8
0.5
0.4
1.0
0.5
0.4
11.1
9.9
9.6
8.9
8.8
4.6
3.9
3.9
2.9
2.6
2.1
MCM-41
MCM-41-Cl
MCM-41-Pr
There are some trends seen in the data in Table 2. Com-
parison between SiO -Pr (L) and SiO -Pr (H) suggests there
2 2
may be a link between the surface area of the support and
the level of functionalisation. The fact that the much higher
surface area of MCM-41 does not correspond to a proportion-
ally higher level of functionalisation may reflect a less acces-
sible surface in the deep pores in this material, or it may be
a consequence of a less reactive surface arising from high
temperature calcination and a relatively low concentration of
reactive surface silanol groups.
The extent to which chloropropyl groups are reacted with
the proline derivatives is also significant. This is quantitative
for SiO
2
-Cl (H) but only about 70% conversion occurs for
SiO -Cl (L) and MCM-41-Cl. It is possible that this is related
2
Fig. 4 TGA-curve for SiO
2
-Cl (H).
to the facility with which 4-hydroxyproline can access the sur-
face chloropropyl groups, although there is no obvious link
with average pore diameter.
The data in Table 2 for the functionalised silicas prepared
by the protection/deprotection routes show that the act of
protecting and then de-protecting results in significant losses
of functional group concentrations, especially through the
esterification of the carboxylic acid group and its subsequent
hydrolysis.
the catalysts. There are two weight loss steps, at 20–200 °C
and 200–500 °C. The first is related to the drying of silica.
The second, which is the more important, is attributed to the
decomposition of the organic groups grafted on the surface
(
after correction for weight loss over the same temperature
range for the parent silica support). Elemental analysis data
for nitrogen and for chlorine (Table 2) are taken as indicators
of the molar concentrations of the supported proline and the
unreacted chloropropyl groups on the silica surface. These
values are checked for their consistency with TGA data. Using
the ratio of the concentrations of the supported proline and
The FTIR spectrum of SiO -Pr (L) is shown in Fig. 5, as an
example of a catalyst prepared without the use of protecting
2
−
1
groups. The band at 2970 cm is characteristic of CH
2
stretch
−
1
of the propyl chain. The vibrations at 1058 cm (Si–O–Si
asymmetric stretching), 799 cm (symmetric stretching),
452 cm (bending) and 957 cm (Si–OH) are characteristic of
−
1
−
1
−1
Table 2 TGA and elemental analytical for the synthesized materials
a
−1
b
−1
b
−1
Sample
Organic content /mmol g
N content /mmol g
Cl content /mmol g
SiO
SiO
SiO
SiO
SiO
SiO
SiO
SiO
2
2
2
2
2
2
2
2
-Cl (L)
-Pr (L)
-Pr-Boc (L)
-Pr-Boc,Me IJL)
-Pr (Boc)deprotected (L)
-Pr (Boc,Me)deprotected (L)
-Cl (H)
0.62
0.61
0.63
0.61
0.58
0.33
1.31
1.30
1.12
1.10
—
0.64
0.20
0.21
0.22
0.19
0.14
1.32
—
0.43
0.42
0.39
0.41
0.17
—
1.30
—
0.71
-Pr (H)
MCM-41-Cl
MCM-41-Pr
1.05
0.29
a
b
Thermogravimetric data. Elemental analysis data.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
for the supported proline catalysts prepared without using
the protection/deprotection routes (entries 3, 4 and 5), and
using the protecting groups (entries 6 and 7). Data for
4
-hydroxy-L-proline in homogeneous solution (at elevated
concentration) (entry 8) is shown, and data for the same
homogeneous 4-hydroxy-proline catalyst with protection
of the amine, N-Boc-trans-4-hydroxy-L-proline (entry 9), is
also given.
2
Entry 10 shows that SiO -Pr (H) is readily recycled and
loses negligible activity and negligible enantioselectivity over
four reaction cycles. Leaching tests were also performed as
described above and no evidence for catalyst leaching to the
reaction mixture was found.
2
Fig. 5 FTIR spectrum of SiO -Pr (L).
MCM-41-Pr, SiO -Pr (H) and SiO -Pr (L) catalyse this trans-
2
2
formation effectively (entries 3, 4 and 5). Overall activities are
higher than, or comparable with, those of the catalysts pre-
pared by the protection/deprotection routes. Turnover fre-
quencies (based on nitrogen content) are lower however.
Enantiomeric excesses are recorded with all three, although
they are not consistent. A possible explanation for the wide
differences might be linked to the involvement of surface
silanol groups on the catalysts. The enantioselectivity of pro-
line is believed to arise from a relatively rigid transition state
involving coordination of a proline-derived enamine and the
aldehyde via hydrogen bonding to the carboxylic acid group.
It is possible that acidic silanol groups on the silica surface
interfere, competing with the carboxylic acid groups for the
localisation of the aldehyde, and the extent of this competi-
tion might reasonably be expected to differ between supports.
The catalysts prepared by the protection/deprotection
routes (entries 6 and 7) show moderate activity, in line with
other reports but with higher TOFs than for those prepared
by the simpler route. It is significant that turnover frequen-
cies are the same for the catalyst prepared by protection of
both amine and carboxylic acid groups (entry 7) and the cata-
lyst where only the amine is protected (entry 6). The former
catalyst shows very low enantioselectivity which possibly sug-
gests that the ester group used to protect the carboxylic acid
is difficult to remove since, as discussed above, enantio-
selectivity is thought to be controlled by an intermediate
involving the free acid group.
the silica support. The very weak bands at 3273, 3145 and
−
1
1
570 cm can be assigned to N–H stretching, hydrogen-
bonded N–H stretching and the carbonyl group (carboxylic
2
8–30
acid) respectively.
These three weak vibrational bands
are characteristic of the secondary amine and carboxylic acid
functional groups on proline. Their low intensity is a conse-
quence of the relatively low concentration of proline on the
silica, but their presence confirms that the proline functional
groups have been largely unaffected through reaction with
surface tethers.
Catalytic data
The catalytic data for the supported organocatalysts in the
aldol reaction are summarised in Table 3.
Detailed kinetic data is given in the ESI.† In all cases, the
aldol reactions were found to follow pseudo-first order kinet-
ics (first order in nitrobenzaldehyde) up to and beyond the
1
7
6
h period. Turnover frequencies (TOFs) are based on nitro-
benzaldehyde conversion and first order plots, assuming that
the supported proline concentration is given by the nitrogen
content (Table 2). In most cases a small amount of the aldol
dehydration product, 4-IJ4-nitrophenyl)but-3-en-2-one (“C” in
ESI†), was also detected.
Data for MCM-41 and MCM-41-Cl (entries 1 and 2) show
that the support materials exhibit no activity. Data is shown
Table 3 Catalytic data for the aldol condensation reaction between 4-nitrobenzaldehyde and acetone using proline catalysts
a
−1
a
b
Catalyst
Conversion of 1/%
TOF/h
Yield of 2/%
ee /%
1
2
3
4
5
6
7
8
9
1
MCM-41
MCM-41-Cl
MCM-41-Pr
0
0
0
0
0
0
—
—
27
63
51
37
2
37
—
65
44
42
23
35
17
35
0
2.7
1.4
2.0
3.5
3.6
0.4
0
41
40
22
34
12
35
0
SiO
SiO
SiO
SiO
2
2
2
2
-Pr (H)
-Pr (L)
-Pr (Boc)deprotected (L)
-Pr (Boc,Me)deprotected (L)
c
Hydroxyproline (dissolved)
Hydroxyproline-Bocprotected
d
0
2
SiO -Pr (H)
41
1.4
38
a
Reaction conditions: 4-nitrobenzaldehyde (151 mg, 1.0 mmol), acetone (10 mL), solid catalyst (0.05 g), 50 °C for 6 h. 95% confidence limit =
b
c
d
±
3%. Enantiomeric excess 95% confidence limit = ±2%. 0.19 mmol of hydroxyproline. Re-used three times: activity on fourth cycle.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol.

View Article Online
Paper
Remarkably, 4-hydroxy-L-proline used as a homogeneous
Catalysis Science & Technology
Table 4 Conversion and yield data in the aldol condensation between
-nitrobenzaldehyde and acetone in DMSO at 25 °C for 20 h
4
catalyst (entry 8) shows significantly lower activity than the
supported proline catalysts (the conversion data corresponds
to the use of much more 4-hydroxy-L-proline in homogeneous
solution than was used on the supported catalysts). The use
of 4-hydroxy-L-proline as a homogeneous analogue to the
supported proline catalysts, rather than L-proline itself, is jus-
tified on the basis that it has been shown by others to be a
−1
Catalyst
Conversion of 1/% TOF/h
Yield of 2/% ee/%
MCM-41-Pr
SiO
SiO -Pr (L)
Hydroxyproline 67
74
90
54
1.9
1.8
1.8
0.3
72
87
51
65
77
75
59
78
2
-Pr (H)
2
1
1
more active catalyst than L-proline in typical aldol reactions.
Table 5 Conversion and yield data in the aldol condensation between
acetone and a range of aldehydes in DMSO at 25 °C for 20 h, using
MCM-41-Pr catalyst
We can therefore be sure that homogeneous L-proline would
also be very much less active than the supported proline cata-
lysts. The higher activities of supported proline suggest that
the silica support enhances the activity of tethered proline,
possibly through interaction between the carboxylic acid
group on the proline and nearby silanol groups, as men-
Catalyst
-Nitrobenzaldehyde
Conversion of 1/% Yield of 2/% ee/%
4
74
72
43
69
56
77
70
82
73
4-Isopropylbenzaldehyde 47
3
1
Benzaldehyde
-Chlorobenzaldehyde
69
59
tioned briefly above.
4
The protected homogeneous catalyst N-Boc-trans-4-
hydroxy-L-proline (entry 9) shows no significant activity,
confirming that the secondary amine is required to catalyse
the aldol reaction. This observation is in agreement with
Supported proline catalysts in
dimethyl sulfoxide (DMSO)
7
,11,18
reported work.
In conclusion, it seems that it is possible to prepare an
effective supported chiral proline catalyst without having to
protect the amine and carboxylic acid groups through the
tethering process. The catalytic and characterisation data sug-
gest that the hydroxyproline reacts with the chloropropyl
tether via the hydroxy group (thereby leaving free the second-
ary amine and carboxylic acid functionalities), Higher levels
of proline functionalisation are achieved when the protec-
tion/deprotection steps are avoided, resulting in higher over-
all activities. However, the lower nominal TOFs observed for
these catalysts may mean that not all the supported proline
is in an active form, and is possibly bound to the surface in
some way other than through reaction between the hydroxy
and chloropropyl groups.
Proline-catalyzed aldol reactions are typically carried out in
organic solvents such as dimethyl sulfoxide (DMSO),
dimethylformamide (DMF) or chloroform. Dimethyl sulfoxide
2
4
has been identified as the solvent of choice in many cases.
In the work reported above, the reactant acetone was also the
solvent. We also carried out catalytic experiments in DMSO/
acetone (4/1 v/v) mixtures at 25 °C. Kinetic data and enantio-
meric excesses for these reactions are shown in Table 4. Bear-
ing in mind the much lower reaction temperature, it is clear
that activities are higher in this solvent system than in ace-
3
3
tone. This is consistent with DFT calculations
which
have shown that the energy barrier to complexation between
−
1
acetone and proline is 171.4 kJ mol , but this is reduced to
−
1
4
0.7 kJ mol in the presence of excess DMSO. The supported
We admit that it is at first surprising that the tethering
process works through reaction between the chloropropyl
group on the silica surface and the hydroxy group on proline,
and not the secondary amine. We suggest that this may be
because the mildly acidic silanol groups on the silica surface
protonate the amine, effectively protecting it during reaction
proline catalysts are more active than hydroxyproline in
homogeneous solution, as in acetone. Enantioselectivities are
also higher, but this may be simply a consequence of the
lower reaction temperature.
Supported proline catalysts with
other aromatic aldehydes
3
2
(
a
the pK of silanol groups can be as low as 3). The proton-
ated amine group would also serve as protection against any
racemisation at the α-carbon atom, a major concern in
amino acid chemistry.
To examine the generality of the reaction, a variety of aro-
matic aldehydes were also used in the reaction. As shown in
Table 5, aromatic aldehydes bearing electron-withdrawing
groups, such as chloro- and nitro-groups, undergo aldol con-
densation smoothly in the presence of MCM-41-Pr in DMSO/
acetone (v/v: 4/1). The reactions generate the corresponding
aldols in good yields (59–74%) and high enantioselectivities
2
9
The supported proline is clearly more active towards aldol
condensation than dissolved 4-hydroxy-L-proline and, by
1
1
extension, proline itself. There appears to be some coopera-
tive catalytic mechanism through which the support
enhances activity. This could be because the catalytic amine
group is activated by hydrogen-bonding to adjacent surface
silanol groups, or perhaps these same silanol groups could
play a role in activating the reactant ketone or aldehyde, or
the intermediate enamine. Finally, it may be that the silanol
groups simply act to concentrate the reactants by surface
adsorption.
(
73–77% ee).
Conclusions
L-Proline-functionalized mesoporous silica has been success-
fully synthesized by a relatively simple route. The effect of
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
protecting the amine and the carboxylic acid groups through
the synthesis of the supported proline on the ultimate activity
and enantiomeric selectivity of the catalyst has been investi-
gated. The results show that the most active and selective cat-
alysts are prepared without protecting the two functional
groups on the proline. This new route could be an economi-
cal and green step towards chiral catalysts in general and
solid-supported proline catalysts in particular.
14 M. Benaglia, G. Celentano and F. Cozzi, Adv. Synth. Catal.,
2001, 343, 171–173.
15 M. Benaglia, M. Cinquini, F. Cozzi, A. Puglisi and
G. Celentano, Adv. Synth. Catal., 2002, 344, 533–542.
16 M. Benaglia, A. Puglisi and F. Cozzi, Chem. Rev., 2003, 103,
3401–3430.
17 D. Dhar, I. Beadham and S. Chandrasekaran, J. Chem. Sci.,
2003, 115, 365–372.
Furthermore, there seems to be a relationship between
the nature of the silica support and the ultimate activity and
enantioselectivity of the supported proline. It seems that a
high surface area for the support is beneficial and it is possi-
ble that pore size is also important in terms of imparting
enantioselectivity to the catalyst. Current research is directed
at ordered mesoporous silica supports with a range of pore
sizes, in an effort to identify the relationship between the
porosity of the support material and the activity and selectivity
of supported proline catalysts.
18 F. Calderón, R. Fernández, F. Sánchez and A. Fernández-Mayoralas,
Adv. Synth. Catal., 2005, 347, 1395–1403.
19 L.-H. Hsiao, S.-Y. Chen, S.-J. Huang, S.-B. Liu, P.-H. Chen,
J. C. C. Chan and S. Cheng, Appl. Catal., A, 2009, 359,
96–107.
20 K. Arya, U. C. Rajesh and D. S. Rawat, Green Chem., 2012, 14,
3344–3351.
21 T. E. Kristensen, K. Vestli, K. A. Fredriksen, F. K. Hansen
and T. Hansen, Org. Lett., 2009, 11, 2968–2971.
22 A. A. Elmekawy, N. R. Shiju, G. Rothenberg and D. R. Brown,
Ind. Eng. Chem. Res., 2014, DOI: 10.1021/ie500839m.
2
3 M. R. Mello, D. Phanon, G. Q. Silveira, P. L. Llewellyn and
C. M. Ronconi, Microporous Mesoporous Mater., 2011, 143,
174–179.
Notes and references
1
2
R. Zawirska-Wojtasiak, Food Chem., 2004, 86, 113–118.
G. Flores, G. P. Blanch and M. L. Ruiz del Castillo, Food
Chem., 2013, 141, 2982–2987.
24 K. K. Sharma, A. V. Biradar, S. Das and T. Asefa, Eur. J.
Inorg. Chem., 2011, 2011, 3174–3182.
3
4
5
M. Kauranen, T. Verbiest, J. J. Maki and A. Persoons, Synth.
Met., 1996, 81, 117–120.
J. Lu, L. Wu, L. Jing, X. Xu and X. Zhang, Dyes Pigm., 2012,
25 J. Qi, B. Qin, J. Liu, Y. Yu, Z. Zhang, W. Zhang, Q. Cai and
W. Zhu, CrystEngComm, 2011, 13, 4666–4675.
26 M. M. J. A. K. M. Lassen, ARKIVOC, 2010, viii, 189–250.
27 P. Iliade, I. Miletto, S. Coluccia and G. Berlier, Res. Chem.
Intermed., 2012, 38, 785–794.
28 W. He, F. Zhang, X. Shi and H. Li, Eur. J. Org. Chem., 2012,
2012, 3753–3758.
29 Z. An, W. Zhang, H. Shi and J. He, J. Catal., 2006, 241,
319–327.
9
4, 169–174.
R. Celis, B. Gámiz, M. A. Adelino, M. C. Hermosín and
J. Cornejo, Sci. Total Environ., 2013, 444, 288–297.
R. N. Patel, Coord. Chem. Rev., 2008, 252, 659–701.
6
7
K. Schulz, L. Ratjen and J. Martens, Tetrahedron, 2011, 67,
5
46–553.
8
9
R. Noyori, Angew. Chem., Int. Ed., 2002, 41, 2008–2022.
U. Eder, G. Sauer and R. Wiechert, Angew. Chem., Int. Ed.
Engl., 1971, 10, 496–497.
30 H. Yang, S. Li, X. Wang, F. Zhang, X. Zhong, Z. Dong
and J. Ma, J. Mol. Catal. A: Chem., 2012, 363–364,
404–410.
1
1
0 Z. G. Hajos and D. R. Parrish, J. Org. Chem., 1974, 39, 1615–1621.
1 B. List, R. A. Lerner and C. F. Barbas, J. Am. Chem. Soc.,
31 A. Zamboulis, N. J. Rahier, M. Gehringer, X. Cattoën, G. Niel,
C. Bied, J. J. E. Moreau and M. W. C. Man, Tetrahedron:
Asymmetry, 2009, 20, 2880–2885.
2
000, 122, 2395–2396.
1
1
2 G. Sabitha, N. Fatima, E. V. Reddy and J. S. Yadav, Adv.
Synth. Catal., 2005, 347, 1353–1355.
3 N. S. Chowdari, J. T. Suri and C. F. Barbas, Org. Lett., 2004,
32 K. Leung, I. M. B. Nielsen and L. J. Criscenti, J. Am. Chem.
Soc., 2009, 131, 18358–18365.
33 K. N. Rankin, J. W. Gauld and R. J. Boyd, J. Phys. Chem. A,
2002, 106, 5155–5159.
6
, 2507–2510.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol.