Chiral hybrid inorganic-organic materials: Synthesis, characterization, and application in stereoselective organocatalytic cycloadditions

Article abstract of DOI:10.1021/jo401852v

The synthesis of chiral imidazolidinones on mesoporous silica nanoparticles, exploiting two different anchoring sites and two different linkers, is reported. Catalysts 1-4 were prepared starting from l-phenylalanine or l-tyrosine methyl esters and support

Full text of DOI:10.1021/jo401852v

Article
pubs.acs.org/joc
Chiral Hybrid Inorganic−Organic Materials: Synthesis,
Characterization, and Application in Stereoselective
Organocatalytic Cycloadditions
Alessandra Puglisi,* Maurizio Benaglia, Rita Annunziata, Valerio Chiroli, Riccardo Porta,
and Antonella Gervasini
̀
Dipartimento di Chimica, Universita degli Studi di Milano, Via Golgi 19, 20133, Milano, Italy
S
* Supporting Information
ABSTRACT: The synthesis of chiral imidazolidinones on mesoporous silica
nanoparticles, exploiting two different anchoring sites and two different
linkers, is reported. Catalysts 1−4 were prepared starting from L-phenylalanine
or L-tyrosine methyl esters and supporting the imidazolidinone onto silica by
grafting protocols or azide−alkyne copper(I)-catalyzed cycloaddition. The four
catalysts were fully characterized by solid-state NMR, N2 physisorption, SEM,
and TGA in order to provide structural assessments, including an evaluation
of surface areas, pore dimensions, and catalyst loading. They were used in
organocatalyzed Diels−Alder cycloadditions between cyclopentadiene and
different aldehydes, affording results comparable to those obtained with the
nonsupported catalyst (up to 91% yield and 92% ee in the model reaction
between cyclopentadiene and cinnamic aldehyde). The catalysts were
recovered from the reaction mixture by simple filtration or centrifugation.
The most active catalyst was recycled two times with some loss of catalytic efficiency and a small erosion of ee.
including the attachment of the chiral catalysts to organic
polymers, dendrimers, membrane supports, and porous inorganic
oxides and via biphasic systems.7 Along this line, the
immobilization of a chiral organic catalyst seems particularly
attractive, because the metal-free nature of these compounds
avoids from the outset the problem of the leaching of the metal,
which often negatively affects and practically prevents the efficient
recycling of a supported organometallic catalyst.8 One of the most
popular and versatile classes of chiral metal-free catalysts is that of
chiral imidazolidinones,9 the so-called MacMillan catalysts. These
compounds have been covalently immobilized on both soluble10
and insoluble supports.11 In 2006 a properly modified enantiopure
first-generation MacMillan imidazolidinone was anchored to novel
siliceous mesocellular foams,12 and more recently it has been
immobilized in the pores of silica gel with the aid of an ionic
liquid,13 on hollow periodic mesoporous organosilica spheres14a
and on magnetic nanoparticles.14b Despite this variety of proposed
solutions, however, the development of an easily available,
inexpensive, and truly efficient recoverable MacMillan catalyst
has yet to be realized. Lower enantioselectivities with respect
to that of the nonsupported system and recyclability are issues
still waiting for a satisfactory solution. Because of the high surface
area and the well-ordered structure, mesoporous silica may
be considered as an ideal support for the preparation of
INTRODUCTION
■
With the advent of combinatorial chemistry many solid-phase
methodologies have been developed where the substrates are
immobilized on a solid support while the reagents are passed
through in the mobile phase.1 While after 20 years of intense
activity in the field it is still a matter of discussion whether
medicinal chemistry and the pharmaceutical industry have
really taken advantage of combinatorial methods,2 it is evident
that organic synthesis has somehow been revolutionized since
the importance and the opportunities offered by automation
and advanced technologies became clear to synthetic chemists.3
For example, the need for more and more efficient high-speed
parallel synthesis, largely employed by pharmaceutical companies
for lead discovery and optimization, led to the development of
alternative solution-phase techniques utilizing supported reagents
or scavangers.4 Solid-phase-assisted synthesis is only one of the
so-called enabling technologies, developed with the aim of
speeding up the synthetic transformation and facilitating the
isolation and purification of the final product.5
While immobilized reagents and scavengers have found
widespread application in organic synthesis, the use of
supported catalysts, and especially of chiral catalytic species,
is much less common. Heterogenization can potentially provide
easily recyclable and reusable solid catalysts6 that have evenly
distributed and precisely engineered active sites, similar to those
of their homogeneous counterparts, and therefore could combine
the advantages of both homogeneous and heterogeneous
systems. Many immobilization approaches have been explored,
Received: August 24, 2013
Published: October 17, 2013
© 2013 American Chemical Society
11326
dx.doi.org/10.1021/jo401852v | J. Org. Chem. 2013, 78, 11326−11334

The Journal of Organic Chemistry
Article
Figure 1. MacMillan imidazolidinone and silica-supported imidazolidinones.
Scheme 1. Synthesis of Supported Imidazolidinone 1
heterogeneous catalysts and may help to positively address
those problems.15
RESULTS AND DISCUSSION
■
Synthesis of the Silica-Supported Catalysts. In design-
Basically two synthetic protocols are available to obtain
mesoporous silica-supported catalysts: the cocondensation
method and postsynthetic grafting. In addition, the nature of
the cocondensation process, which involves the reaction of a
tetralkoxysilane with organotrialkoxysilanes carrying the differ-
ent functionalities required for catalysis, can open access to
systems in which the functional group ratios can be varied at
will, allowing an additional tuning of the catalyst’s properties. A
few years ago we developed a bifunctional catalytic system
supported on mesoporous silica featuring tertiary amine and
thiourea functionalities in different ratios, where the coopera-
tivity of the two catalytically active sites was demonstrated in
the conjugate addition of acetylacetone to 2-nitrostyrene.16 On
the basis of our previous experience we decided to investigate
the use of these materials as supports also for chiral organic
catalysts. In this work we report the preparation of a silica-
supported MacMillan catalyst through different immobilization
strategies, the characterization of the functionalized materials,
and the study of their catalytic behavior in stereoselective
Diels−Alder reactions; finally, the recovery and the recyclability
of the supported chiral organocatalysts were also studied.
ing the synthesis of silica-supported chiral imidazolidinones, we
reasoned that, in principle, the most convenient route could
involve the cocondensation of a silica precursor (namely TEOS,
tetraethoxysilane) with a chiral organotrialkoxysilane bearing
the imidazolidinone moiety; however, the drastic reaction
conditions required for the cocondensation, in particular treatment
of the siliceous material with HCl in MeOH at 60 °C, could lead to
extensive degradation of the imidazolidinone ring. We therefore
opted for a postgrafting functionalization of an already synthesized
material with a chiral trialkoxysilane. With regard to the silica matrix,
we chose to use mesoporous silica nanoparticles (MSNs) prepared
from tetraethoxysilane in the presence of cetyltrimethylammonium
bromide (CTAB) as the surfactant employed as templating agent.
In principle, imidazolidin-4-ones such as MacMillan’s first-
generation catalysts offer two different sites for attachment to
the solid support: the amide nitrogen at position 3 and the aryl
residue at the stereogenic center (Figure 1).
Both opportunities were explored in this work: the condensation
of allyl amine with (S)-phenylalanine methyl ester followed by an
acid-catalyzed reaction with acetone afforded imidazolidinone 6
(Scheme 1) in 90% yield after one chromatographic purification.
11327
dx.doi.org/10.1021/jo401852v | J. Org. Chem. 2013, 78, 11326−11334

The Journal of Organic Chemistry
Article
Scheme 2. Synthesis of Supported Imidazolidinone 2
Scheme 3. Synthesis of Supported Azide 13 and Imidazolidinones 3 and 4
Platinum-catalyzed hydrosilylation with trimethoxysilane led to the
enantiopure trialkoxysilane 7, which was grafted to mesoporous
silica in toluene at 60 °C for 24 h to afford catalyst 1.
investigated. Copper-catalyzed azide/alkyne dipolar cycloaddition
was employed to connect the chiral catalyst to the properly
derivatized siliceous material (Scheme 3). Siloxane 12 was readily
prepared from commercially available (3-chloropropyl)-
triethoxysilane and sodium azide in CH3CN and was used in
the cocondensation reaction with TEOS in the presence of
CTAB to afford mesoporous silica-supported azide 13.
Imidazolidinone 9 was reacted with propargyl bromide to afford
derivative 14, which was subjected to copper-catalyzed cyclo-
addition with azide 13 to give catalyst 3. The latter was subjected
to a capping reaction by treatment with hexamethyldisilazane to
give catalyst 4, in order to evaluate if the free silanol groups on
the silica surface could participate in the catalytic cycle by
forming hydrogen bonds with the reagents.
Another option is offered by the use of (S)-tyrosine instead
of (S)-phenylalanine to generate an imidazolidinone already
equipped with a properly located and chemically suitable
handle for the hetereogenization process.17 Thus, starting from
(S)-tyrosine methyl ester hydrochloride, imidazolidinone 9 was
easily obtained in 77% yield by N-butyl amide formation and
treatment with acetone (Scheme 2). Reaction with allyl
bromide in acetonitrile in the presence of Cs2CO3 allowed us
to introduce the carbon−carbon double bond, instrumental for
performing catalyst immobilization by platinum-catalyzed
hydrosilylation. Grafting of the functionalized imidazolidinone
11 onto mesoporous silica nanoparticles (MSNs) in toluene at
60 °C for 24 h afforded the supported catalyst 2.
Preliminary experiments to assess the catalytic behavior of
the immobilized catalysts 1 and 2 in the Diels−Alder cycloaddi-
tion between cinnamic aldehyde and cyclopentadiene18 showed
better performance for the latter (see the discussion of catalytic
experiments); therefore, the study focused on the (S)-tyrosine-
derived imidazolidinone and a different strategy to immobilize it was
Characterization of Supported Catalysts. 13C and 29Si
MAS NMR. Solid-state 29Si NMR experiments have been
extensively used to characterize different surface incorporation
patterns and conformations of siliceous material functionalized
with organic residues. Nuclear magnetic resonance studies of
silica-based materials are potentially useful, where applicable,
since they provide well-resolved spectra that can easily
characterize the bonding environment of silicon atoms near
11328
dx.doi.org/10.1021/jo401852v | J. Org. Chem. 2013, 78, 11326−11334

The Journal of Organic Chemistry
Article
the surface.19−21 In this study, cross-polarization (CP) and
direct-polarization (DP) magic angle spinning (MAS) 13C and
29Si NMR experiments were performed to investigate catalysts
1−4 in order to obtain evidence for the presence of the organic
moieties in the mesoporous material and confirm their chemical
structure. It is well recognized that a given surface silicon atom
bearing a functional group can have a variable number n of Si−O−Si
bonds, leading to so-called T1, T2, and T3 substructures. The bonding
schemes, for these hybrid materials containing organic functionalities,
lead to very different geometries on the surface, as illustrated in
Figure 2. Knowledge of the surface incorporation patterns and
The 29Si solid-state spectra of catalysts 1−3 (Figure 3) are
dominated by the resonance lines associated with the silicon
sites Q4, Q3, and Q2. Moreover, the spectra clearly showed very
different patterns for the functionalized Tn species, where the
silicon atoms are directly bound to at least one organic moiety.
The presence of peaks assigned to T3, T2 ,and T1 showed
that the organic groups are indeed covalently bound to the
surface. Deconvolution analysis of 29Si DP spectra of catalysts
1−3 were performed to recognize Tn and Qn relative
concentrations in order to determine the surface coverage
(SC) and the molar concentrations of organic moieties, ranging
from 0.62 to 1.43 mmol/g (for Tn and Qn determinations, see
the Supporting Information).
Finally, the 13C spectra of catalysts 1−4 demonstrated that
the mesopores were indeed functionalized as expected and the
organic residues were stably bounded to the inorganic material
(see the Supporting Information for details).
Morphological Properties. The morphological examination
of the sample particles was performed by collecting images of
the surfaces by SEM analysis. Figure 4 (left) shows the
micrograph of azide 13, precursor of catalyst 3, at high
magnification. As expected,16 rod-shaped particles of different
lengths and diameters (ca. 0.4 μm in length and 500 nm in
diameter) with curved hexagonal-shaped tubular morphology
completely cover irregular polyhedral-shaped grains, which can
be associated with the silica morphology. Anchoring of the
catalyst to the azido silica 13 did not deeply affect the
morphology of the final particles. As can be seen by a SEM
micrograph of the surface of 3 (Figure 4 right), an even more
dense surface of rod-shaped tubular particles covers the silica
grains which can now only be guessed (for more SEM images
see the Supporting Information).
Figure 2. Possible Si substitution on the silica surface.
the conformations of functional groups for such materials is an
essential step toward developing efficient functional substrates.
The 29Si DPMAS experiments, carried out according to
previously described methods,21 allowed us to perform
quantitative measurements and determine the catalyst loading,
providing the functionalization degree of the SiO2 materials with
the organic siloxane moieties. The 29Si spectra gave also
information about the bulk surface species Q4 [(SiO)4 Si] and
proton-rich Qn sites: Q3 [(SiO)3 SiOH], Q2 [(SiO)2 Si(OH)2],
and Q1 [(SiO) Si(OH)3] (Figure 2).
The textural properties in terms of surface areas, pore
volumes, and pore size distributions of the azide 13 and catalyst
Figure 3. 29Si CPMAS (left column) and DPMAS (rigth column) NMR spectra obtained for mesoporous catalysts 1−3.
11329
dx.doi.org/10.1021/jo401852v | J. Org. Chem. 2013, 78, 11326−11334

The Journal of Organic Chemistry
Article
Figure 4. SEM images of a sample of the azide surface 13 (left) and of catalyst 3 (right) at magnifications of 10000×.
Figure 5. TGA (left) and DTGA (right) analysis of 13 and 3 samples. The linear rate increase (5 °C/min) used in the experiment is indicated on the
right axis.
3 have been determined by N2 adsorption−desorption
isotherms. The two surfaces have characteristic type IV BET
isotherms consistent with the presence of cylindrical mesoscale
pores. The BET surface areas of 13 and 3 are 1109 and
486.5 m2 g−1, respectively. The loss of surface area observed for
catalyst 3 could be due to the higher density of the organic
functionalities loaded on the azide sample which completely
cover the silica surface. The pore size distribution, calculated by
the BJH approach, showed a defined pore population of sizes
around 20 and 16 Å for 13 and 3, respectively (see the
Supporting Information for N2 adsorption−desorption iso-
therms and calculated pore size BJH distribution curves).
In order to have an estimate of the total amount of the
organic residues loaded on the silica surface, thermogravimetric
analysis (TGA) was carried out under a linear increase rate of
temperature (up to 450 °C) and in an air atmosphere to obtain
the oxidative removal of all the organic residues bound to the
silica surface. Figure 5 shows both the thermograms obtained in
terms of residual mass as a function of time and the derivative
of the mass loss, the DTGA curves, for samples 13 and 3.
For both samples 3 and 13 an initial mass loss likely
associated with the desorption of some water and organic
species from the silica surface could be observed. Azide 13
showed two transitions at 246 and 390 °C with a total mass loss
of 10.9% (estimated between 100 and 450 °C temperature
interval); the calculated amount of organic residue removed
was 1.29 mmol/g of silica. Catalyst 3 showed a first transition at
159 °C followed by a very intense transition at 364 °C, with a
very broad and high peak leading to a total mass loss of 21.12%.
Although at the moment there is no single reliable analytical
technique for the quantification of different organic residues on
silica matrices, by using a combination of several methods it is
possible to gain detailed information about functionalized
materials. In particular, well-established analytical techniques
such as MAS NMR could be used not only to characterize
different surface incorporation patterns but also to determine
the ratio between two different organic residues.
Catalytic Behavior of the Immobilized Catalysts.
Catalysts 1−4 were tested in the Diels−Alder reaction between
cinnamic aldehyde (1 equiv) and cyclopentadiene (5 equiv) in
MeOH/H2O or CH3CN/H2O at room temperature for 72 h
using 30 mol % of the supported catalyst.22 HBF4 was added to
the reaction mixture in order to protonate the imidazolidinone
and generate the actual catalytic species.10b The results are
summarized in Table 1.
Catalyst 1 afforded cycloadducts 15 in 74% yield with 72% ee
for the endo isomer and 76% ee for the exo isomer (Table 1,
entry 1); catalyst 2 performed comparably in terms of
stereochemical activity (76% ee for the exo isomer) but much
better in terms of chemical activity, affording the cycloadducts
quantitatively (entry 3). Since preforming the catalyst led to a
marked decrease of both yield and stereoselection (entry 2), the
in situ addition of HBF4 was preferred. Use of trifluoroacetic acid
(TFA) as protonating agent led to improved stereoselectivity
(81% ee for the endo isomer and 78% ee for the exo isomer) and
lower yield (entry 5) in CH3CN/H2O as solvent. Catalysts
derived from the copper-catalyzed azide−alkyne cycloaddition
(catalysts 3 and 4) performed better than their grafted
11330
dx.doi.org/10.1021/jo401852v | J. Org. Chem. 2013, 78, 11326−11334

The Journal of Organic Chemistry
Article
Table 1. Diels−Alder Reactions Promoted by Catalysts 1−4
Table 2. Diels−Alder Reaction between Cyclopentadiene
and Different Aldehydes
a
b
c
entry
R
yield (%)
endo:exo
endo, exo ee (%)
yield
a
endo, exo ee
b
c
entry catalyst
solvent
(%)
endo:exo
(%)
1
2
3
4
CH3
98
84
92
92
41:59
50:50
50:50
45:55
78, 76
88, 85
91, 87
1
2
3
4
5
6
7
8
9
1
1
2
2
2
3
3
4
3
3
MeOH/H2O 95/5
MeOH/H2O 95/5
MeOH/H2O 95/5
CH3CN/H2O 95/5
CH3CN/H2O 95/5
MeOH/H2O 95/5
MeOH/H2O 95/5
MeOH/H2O 95/5
CH3CN/H2O 95/5
CH3CN/H2O 95/5
74
27
98
84
70
66
98
66
75
91
56:44
53:47
28:72
55:45
55:45
56:44
56:44
56:44
55:45
72, 76
42, 61
58, 76
66, 70
81, 78
88, 90
87, 80
84, 91
91, 89
4-Br-Ph
d
4-NO2-Ph
2-NO2-Ph
91, 84
a
b
1
Isolated yield after chromatographic purification. Evaluated by H
NMR on the crude reaction mixture. Evaluated by HPLC or GC on a
chiral stationary phase after reduction to the corresponding alcohol.
e
e
c
One of the main goals of supporting a catalyst onto a solid
matrix is to facilitate its recovery and recyclability. The anchoring
of an organic catalyst onto silica offers the opportunity of
recovering the catalytic species by simple separation from the
crude reaction mixture using filtration or centrifugation. After a
Diels−Alder reaction between cyclopentadiene and cinnamic
aldehyde, catalyst 3 was recovered by centrifugation, washed with
CH2Cl2/MeOH, dried for 3 h at 60 °C under high vacuum, and
reused in the next reaction after activation with HBF4. Results of
this iterative procedure are reported in Table 3.
f
10
55:45
92, 90
a
b
1
Isolated yield after chromatographic purification. Evaluated by H
c
NMR on the crude reaction mixture. Evaluated by HPLC and GC on
a chiral stationary phase after reduction to the corresponding alcohol.
The protonated catalyst was isolated and then used in the reaction.
Trifluoroacetic acid was used. The reaction was set using 5 equiv of
cyclopentadiene; after 20 h, 5 equiv of cyclopentadiene was added for
a total reaction time of 40 h.
d
e
f
counterparts. Catalyst 3 afforded the cycloadducts in 66% yield
with 88% ee for the endo isomer and 90% ee for the exo isomer
(entry 6), results comparable to those obtained by the
nonsupported chiral imidazolidinone;18 using 3/TFA resulted
in an increase of the chemical yield but an erosion of the
enantiomeric excess (entry 7 vs entry 6). Capped catalyst 4
performed equally well in comparison to catalyst 3 (entry 8),
giving the products with 66% yield and 84% ee for the endo
isomer and 91% ee for the exo isomer, pointing out that the free
silanol groups on the silica surface did not alter the catalytic
efficiency of the imidazolidinone. Using catalyst 3/HBF4 in
CH3CN/H2O as the solvent system allowed us to improve both
yield and stereoselection, affording the products in 75% yield and
91% ee for the endo isomer and 89% ee for the exo isomer (entry
9 vs entry 6). By slightly changing the reaction conditions, that is by
adding 5 equiv more of cyclopentadiene after 20 h, we were able to
improve the catalytic efficiency and shorten the reaction time,
obtaining the cycloadducts in quantitative yield, with 92% ee for the
endo isomer and 90% ee for the exo isomer in 40 h (entry 10).
Having thus established the best reaction conditions
(entry 10), we extended the scope of the Diels−Alder reaction
to differently substituted aldehydes. The results are summarized
in Table 2.
Reaction with an aliphatic aldehyde such as crotonic
aldehyde afforded the cycloadducts in quantitative yield with
78% ee for the endo isomer and 76% ee for the exo isomer
(Table 2, entry 1). 4-Bromocinnamic aldehyde reacted with
cyclopentadiene to afford the products in high yield (94%) and
high enantiomeric excess (88% and 85% for the endo and exo
isomers, respectively) (entry 2). Catalyst 3/HBF4 promoted
the reaction between electron-deficient 4-nitrocinnamic
aldehyde and cyclopentadiene with 92% yield and 91% ee for
the endo isomer and 87% ee for the exo isomer (entry 3) and
also between the more sterically demanding 2-nitrocinnamic
aldehyde and cyclopentadiene, affording the cycloadducts with
92% yield and 91% ee for the endo isomer and 84% ee for the
exo isomer (entry 4).
Table 3. Recycling Experiments of Catalyst 3
a
b
c
entry
yield (%)
endo:exo
endo, exo ee (%)
1
2
3
91
63
41
55:45
55:45
55:45
92, 90
87, 83
72, 74
a
b
1
Isolated yield after chromatographic purification. Evaluated by H
c
NMR on the crude reaction mixture. Evaluated by HPLC on a chiral
stationary phase after reduction to the corresponding alcohol.
The cycloadducts were obtained in decreasing yields (63%
and 41% in the second and third runs, respectively) and
decreasing enantiomeric excesses (87% and 72% for the endo
isomer in the second and third runs, respectively). When the
catalyst was employed in a fourth run, no product was obtained.23
While the diminished chemical activity was somehow predictable,
although more recycles were usually possible with different
supported imidazolidinones, the loss of stereochemical activity was
more surprising.10 Table 4 reports some recycling data of different
supported chiral imidazolidinones in the Diels−Alder cyclo-
addition between cyclopentadiene and α,β-unsaturated aldehydes.
Poly(ethylene glycol)-supported imidazolidinone 16 (Figure 6)
could be recovered from the reaction mixture between cyclo-
pentadiene and acrolein (67% yield, 92% ee; entry 3, Table 4) and
reused in a second cycle to afford the product in 61% yield, and
87% ee; iteration of the recovery/recycling protocol was possible
for two more cycles, both occurring with almost unchanged
stereoselectivities (87% and 85% ee, respectively) but decreasing
yields (50% and 38%, respectively). When the catalyst was used in
a fifth cycle, the yield was very low (10−15% after 60 h).10a More
recently, our group reported the use of poly(methylhydrosiloxane)-
supported chiral imidazolinone 17, which could be reused five
times in the Diels−Alder cycloaddition between cyclopentadiene
and cinnamic aldehyde with only marginal loss of chemical activity
(yield from 65% to 53%, entries 5 and 6) and with no appreciable
decrease of stereo- and enantiocontrol, affording both exo and endo
11331
dx.doi.org/10.1021/jo401852v | J. Org. Chem. 2013, 78, 11326−11334

The Journal of Organic Chemistry
Article
TGA in order to evaluate structural and morphological
properties. Catalysts 1−4 were tested in the Diels−Alder
cycloaddition between cyclopentadiene and cinnamic aldehyde;
catalyst 3 proved to be the most efficient, affording the product
in yield and enantiomeric excess comparable to those obtained
by using nonsupported chiral imidazolidinones. The scope of
the reaction was also extended to different substituted
aldehydes. Attempts to recover and recycle the catalyst showed
some limitations of the system but offered the opportunity to
get some insights into the stability of the silica network.
Table 4. Comparison of Supported Catalysts 3 and 16−18 in
the Synthesis of Adducts 15
entry
catalyst
support
MSNs
run
yield (%)
endo ee (%)
1
2
3/HBF4
1
3
1
4
1
6
1
8
91
41
67
38
65
53
98
86
92
72
92
85
93
91
81
78
a
3
a
16/TFA
17/HBF4
18/TFA
PEG
4
5
6
7
8
PMHS
PMO
EXPERIMENTAL SECTION
Materials. Dichloro(dicyclopentadienyl)platinum(II) (dcpPtCl2;
Aldrich 97%) was purchased from Sigma-Aldrich and used without
further purification.
■
a
Acrolein instead of cinnamic aldehyde was used.
isomers with enantioselectivity always higher than 90%.10b Finally,
Wang and co-workers showed that hollow periodic mesoporous
organosilica spheres functionalized with MacMillan imidazolidinone
18 could be reused for at least seven runs in the Diels−Alder
cycloaddition between cyclopentadiene and cinnamic aldehyde
without a noticeable decrease in enantioselectivity (from 81% ee to
78% ee, entries 7 and 8), although with some loss of the reaction
yield (from 98% to 86%).14
We reasoned that the difficulty in recovering and recycling
catalyst 3 could be ascribed not only to some degradation of the
imidazolidinone moiety (as demonstrated elsewhere17) but also
to degradation of the silica network, which can be sensitive to
mechanical stress. It seems possible that extensive manipu-
lations of the catalyst, such as the repeated cycles of washing,
stirring, centrifugation, and filtration could ultimately lead to a
collapse of the organic−inorganic silica network, thus limiting
the recyclability of the catalyst. Since we have no evidence at
the moment for supporting our hypothesis, further studies are
currently underway toward a better understanding of the
mechanical properties of the MSNs; in order to avoid grinding
of the material, the use of an orbital shaker during the catalytic
test instead of the common magnetic stirring is pro-
grammed.24
All reactions were carried out in oven-dried glassware with magnetic
stirring under a nitrogen atmosphere, unless otherwise stated. Dry
solvents were purchased and stored under nitrogen over molecular
sieves (bottles with crown caps). Reactions were monitored by
analytical thin-layer chromatography (TLC) using silica gel 60 F254
precoated glass plates (0.25 mm thickness) and visualized using UV
light or phosphomolybdic acid.
NMR Spectroscopy. The 1H and 13C NMR spectra were recorded
at 500.0 and 125.62 MHz, respectively. The spectra were measured
with CDCl3 as solvent and the chemical shifts externally referenced to
TMS.
Solid-State NMR. The 13C and 29Si solid-state NMR spectra were
recorded at 125.62 and 99.36 MHz, respectively, on a Bruker Avance
500 spectrometer, equipped with a 4 mm magic angle spinning (MAS)
broad-band probe (spinning rate vR up to 13 kHz). The MAS spectra
were performed on solid samples, (typically 0.12 g); each sample was
packed into a 4 mm MAS rotor (50 μL sample volume) spinning at
13 kHz at a temperature of 300 K. Direct polarization (DP) and
variable amplitude cross-polarization (CP) methods (contact time
τc 1 ms) were used for recording the 29Si spectra with 400 scans and
a delay of 300.0 and 20000 scans and a delay of 3.0 s, respectively.
The 13C experiments were performed with the CP method using a
contact time τc of 1.5 ms, 70000 scans, and 3.0 s of delay. All
chemical shifts were externally referenced to TMS.
The 29Si resonances were assigned following previous reports in
the literature,21 while those of 13C were referenced to the chemical
shifts found in the solution spectra of the corresponding unbound
compounds.
CONCLUSIONS
■
The synthesis of four imidazolidinones supported on
mesoporous silica nanoparticles was reported. The catalysts
were extensively characterized by different analytical techni-
ques: 29Si and 13C MAS NMR for structural confirmation and
silicon atom substitution, and N2 physisorption, SEM, and
N2 Physisorption. Sample surface area (m2 g−1), pore volume
(cm3 g−1), and pore size distribution were determined from N2
adsorption and desorption isotherms measured at liquid nitrogen
temperature with an automatic surface area analyzer; the BET and
BJH methods were employed for the computations. Prior to the
Figure 6. Supported chiral imidazolidinones.
11332
dx.doi.org/10.1021/jo401852v | J. Org. Chem. 2013, 78, 11326−11334

The Journal of Organic Chemistry
Article
measurement, the sample (ca. 1 g), crushed and sieved as 0.250−
0.354 mm particles, was thermally activated at 90 °C for 4 h under
vacuum. The pore volume was determined from the total amount of
N2 adsorbed at P/P0 = 0.95 using the N2 density in the normal liquid
state (ρ = 0.8081 g cm−3); the N2 molecular area was taken as 16.2 Å2.
The pore size distribution was obtained by the BJH (Barrer, Joyner,
and Halenda) model equation on the basis of the desorption branch of
the N2 isothem.
Thermogravimetry. Thermogravimetric analysis (TGA) was
performed on dried samples in the temperature interval from 50 to
450 °C under flowing air with a ramp of 5 °C min−1 and maintaining
the final temperature for 30 min.
Scanning Electron Microscopy. Scanning electron micrographs
(SEM) were obtained by a FE-SEM instrument operating at 0.2−30 kV.
The powder samples were coated with gold before analysis.
Compounds 5, 6, 8−10, 12, and 14 were prepared according to
published procedures.10,17,25
Compound 13. This compound was prepared according to the
general procedure for the synthesis of MSNs by co-condensation of
azide 12 with TEOS.
General Procedure for Grafting Reaction (Catalysts 1 and 2).
MSNs (1 g) and the properly modified organotrimethoxysilane
(2 mmol) were suspended in toluene (10 mL) and stirred under an inert
atmosphere at reflux for 48 h. The catalyst was isolated by centrifugation,
washed with CH2Cl2 (2 × 10 mL), and dried under vacuum.
Catalyst 3. To a solution of compound 14 (1.47 g, 4.68 mmol) in
tetrahydrofuran (40 mL) were sequentially added N,N-diisopropyle-
thylamine (8.2 mL, 46.8 mmol), copper(I) iodide (10 mg,
0.05 mmol), and azide 13 (3 g), in that order. The mixture was
stirred at 35 °C for 40 h, and then it was centrifuged and the residue
was washed three times with a dichloromethane/methanol mixture
(9/1, v/v).
After centrifugation the solid was recovered and dried under high
vacuum to constant weight.
Catalyst 4. To a suspension of catalyst 3 (1 g) in toluene (6 mL)
was added hexamethyldisilazane (0.42 mL), and the mixture was
stirred at 85 °C. After 18 h it was filtered and the solid residue was
washed sequentially with toluene, methanol, and acetone and dried
under high vacuum to constant weight.
Compound 7. To a solution of imidazolidinone 6 (2.84 g, 11.7 mmol)
in THF (30 mL) were added trimethoxysilane (4.5 mL, 35 mmol) and
dcpPtCl2 (48 mg 0.12 mmol), in that order. The mixture was refluxed
under nitrogen for 24 h; the solvent was evaporated, and the crude
imidazolidinone 7 was obtained in quantitative yield as a brown oil and
General Procedure for Diels−Alder Cycloaddition between
Cyclopentadiene and Cinnamic Aldehyde (Compound 15). To
a suspension of the catalyst (30 mol %) in the solvent mixture (2 mL
of 95/5 acetonitrile/water, v/v) was added tetrafluoroboric acid
solution (48 wt % in water, stoichiometric to the catalyst), and the
mixture was stirred for 10 min. Freshly distilled trans-cinnamic aldehyde
(0.26 mmol, 33 μL) and cyclopentadiene (1.3 mmol, 107 μL) were
added sequentially, and the mixture was stirred at room temperature for
20 h. Freshly distilled cyclopentadiene (1.3 mmol, 107 μL) was added
again, and the mixture was stirred for a further 20 h. After this reaction
time, the mixture was centrifuged and the residue was washed with
dichloromethane (2 × 10 mL). After centrifugation the heterogeneous
catalyst was recovered and dried under high vacuum at 60 °C for 3 h.
The supernatants were concentrated under vacuum, and 1H NMR of the
crude product was performed in order to evaluate the conversion and
the endo:exo ratio. The endo:exo ratio was determined by using the
CHO signals at δ 9.60 (endo) and 9.93 (exo) ppm.
If it was clean enough, the crude product (yellow oil) was dissolved
in 2 mL of methanol and reduced with NaBH4 (2 equiv). The mixture
was stirred at room temperature for 1 h, and then 1 mL of distilled
water was added. The mixture was extracted with dichloromethane
(3 × 5 mL) and then concentrated under vacuum to afford the
alcohol. This product is known and was purified by flash column
chromatography on silica gel with a 90/10 hexane/ethyl acetate
mixture as eluent, affording a mixture of endo and exo Diels−Alder
adducts.
Data for the endo product are as follows. Rf = 0.42 (hex/EtOAc 9/1
stained blue with phosphomolybdic acid). 1H NMR (300 MHz,
CDCl3): δ 9.61 (d, J = 2.2 Hz, 1H), 7.14−7.34 (m, 5H), 6.43 (dd, J =
3.3, 5.6 Hz, 1H), 6.18 (dd, J = 2.8, 5.7 Hz, 1H), 3.34 (brs, 1H), 3.14
(brs, 1H), 3.10 (d, J = 4.8, 1H), 2.99 (dd, J = 2.7, 5.4, 1H), 1.82 (d, J =
8.7, 1H), 1.61−1.64 (m, 1H). Data for the exo product are as follows.
Rf = 0.42 (hex/EtOAc 9/1 stained blue with phosphomolybdic acid).
1H NMR (300 MHz, CDCl3): δ 9.93 (d, J = 2.0 Hz, 1H), 7.14−7.34 (m,
5H), 6.34 (dd, J = 3.4, 5.5 Hz, 1H), 6.08 (dd, J = 3.0, 5.5 Hz, 1H), 3.73 (t,
J = 3.8, 1H), 3.23 (m, 2H), 2.60 (dd, J = 1.5, 3.4, 1H), 1.61−1.64 (m, 2H).
Synthesis of ((1S,2S,3S,4R)-3-Phenylbicyclo[2.2.1]hept-5-en-
2-yl)methanol and ((1R,2S,3S,4S)-3-Phenylbicyclo[2.2.1]hept-
5-en-2-yl)methanol. Data for an endo/exo mixture are as follows.
1H NMR (300 MHz, CDCl3): δ 7.18−7.31 (m, 10H), 6.34−6.41
(m, 2H), 6.15−6.20 (m, 1H), 5.93−5.98 (m, 1H), 3.86−3.94 (m, 1H),
3.59−3.70 (m, 3H), 3.39 (t, J = 12.8 Hz, 1H), 3.04 (br s, 2H), 2.83−
2.88 (m, 3H), 2.30−2.42 (m, 2H), 2.15−2.18 (m, 2H), 1.76−1.82
(d, J = 8.9, 1H), 1.65−1.68 (d, J = 8.7, 1H). The enantiomeric excess
was determined by chiral HPLC with a Daicel Chiralcel OJ-H column
(eluent 7/3 hex/IPA; 0.8 mL/min flow rate, detection 225 nm):
tR 11.6 min (endo, minor), tR 23.9 min (endo, major), tR 30.3 min
(exo, minor), tR 40.3 min (exo, major).
1
used without further purification. H NMR (500 MHz, CDCl3, 25 °C,
TMS): δ 7.28 (t, 3J(H,H) = 7.5 Hz, 2H), 7.22 (m, 3H), 3.76 (t, 3J(H,H) =
5 Hz, 1H), 3.6 (s, 9H), 3.3 (part B of an AB system, 1H), 3.05 (m, 2H), 2.9
(part A of an AB system, 1H), 1.55 (m, 2H), 1.25 (s, 3H), 1.15 (s, 3H), 0.6
3
ppm (t, J(H,H) = 7.4 Hz, 2H). 13C NMR (125 MHz, CDCl3, 25 °C,
TMS): δ 173.8, 137.0, 129.6, 128.5, 126.8, 76.0, 58.8, 50.5, 42.8, 37.0, 28.0,
26.5, 22.4, 6.6 ppm.
Compound 11. To a solution of imidazolidinone 10 (1.1 g, 3.47 mmol)
in THF (10 mL) were added trimethoxysilane (1.3 mL, 10.4 mmol)
and dcpPtCl2 (12 mg 0.03 mmol), in that order. The mixture was
refluxed under nitrogen for 24 h; the solvent was evaporated, and the
crude imidazolidinone 11 was obtained in quantitative yield as a brown
oil and used without further purification. 1H NMR (500 MHz, CDCl3,
3
3
25 °C, TMS): δ 7.07 (d, J(H,H) = 8.3 Hz, 2H), 6.77 (d, J(H,H) =
8.3 Hz, 2H), 3.85 (t, 3J(H,H) = 7.9 Hz, 2H), 3.66 (dd, 3J(H,H) = 5.5
and 4.6 Hz, 1H), 3.51 (s, 9H), 3.28 (part B of an AB system, 1H), 3.05
3
(AB system, J(H,H) = 11.0, 5.5, and 4.6 Hz, 2H), 2.85 (part A of an
AB system, 1H), 1.83 (qui, 3J(H,H) = 7.9 Hz, 2H), 1.40 (m, 2H), 1.25
3
(m, 2H), 1.22 (s, 3H), 1.10 (s, 3H), 0.87 (t, J(H,H) = 7.9 Hz, 3H),
0.74 ppm (t, J(H,H) = 7.9 Hz, 2H). 13C NMR (125 MHz, CDCl3,
3
25 °C, TMS): δ 173.9, 158.0, 130.6, 128.4, 114.5, 76.0, 69.6, 58.8, 50.4,
40.2, 35.7, 31.3, 27.9, 26.4, 22.4, 20.3, 13.7, 5.2 ppm.
General Procedure for the Synthesis of MSNs. A solution of
cetyltrimethylammonium bromide (365 mg, 1 mmol) in water
(88 mL) and 2 M NaOH (1.3 mL) was mechanically stirred at
550 rpm at 80 °C for 30 min. The stirring speed was decreased to 200 rpm,
and tetraethoxysilane (TEOS, 1.82 mL, 8.16 mmol) was rapidly added.
In the case of a functionalized material the properly modified
organotrialkoxysilane (1.05 mmol) was added with TEOS. After 2 min
of stirring at 200 rpm a precipitate was formed. Stirring at 500−600 rpm
was continued for 2.0 h at 80 °C, and the mixture was filtered while still
hot. The solid was washed with water (150 mL) and MeOH (150 mL)
and dried under high vacuum for 3 h to afford a white material
(794 mg). This was then treated with a solution of concentrated HCl
(0.6 mL) in MeOH (80 mL) under mechanical stirring for 2.5 h at
60 °C in order to remove the surfactant. The cooled mixture was filtered
and the solid washed again with water and MeOH (100 mL each). The
white solid was dried under high vacuum for 3 h at 90 °C and subjected
to an alkaline wash with a saturated Na2CO3 methanol solution
(100 mL/g). After 3 h of stirring at 550 rpm at room temperature the
mixture was filtered and the solid was washed with water and methanol
and dried under high vacuum at 80 °C for 3 h. The material was finally
suspended in distilled water (100 mL/g) and mechanically stirred at
550 rpm at room temperature for 2 h, filtered on a porous septum,
washed with methanol, and dried under high vacuum at 60 °C to
constant weight.
11333
dx.doi.org/10.1021/jo401852v | J. Org. Chem. 2013, 78, 11326−11334

The Journal of Organic Chemistry
Article
(14) (a) Shi, J. Y.; Wang, C. A.; Li, Z. J.; Wang, Q.; Zhang, Y.; Wang,
W. Chem. Eur. J. 2011, 17, 6206. (b) Riente, P.; Yadav, J.; Pericas, M.
A. Org. Lett. 2012, 14, 3668.
(15) (a) Dickschat, A. T.; Behrends, F.; Surmiak, S.; Weiß, M.;
Eckert, H.; Studer, A. Chem. Commun. 2013, 49, 2195. (b) Dickschat,
ASSOCIATED CONTENT
■
S
* Supporting Information
Text, figures, and tables giving 13C and 29Si MAS NMR of catalysts
1−4, BET isotherms and BJH distributions of compounds 3 and
13, NMR and HPLC spectra of the cycloadducts, and SEM
images of bare silica and catalyst 2. This material is available free of
charge via the Internet at http://pubs.acs.org.
A. T.; Behrends, F.; Buhner, M.; Ren, J.; Weiß, M.; Eckert, H.; Studer,
̈
A. Chem. Eur. J. 2012, 18, 16689. (c) Nakazawa, J.; Smith, B. J.; Stack,
T. D. P. J. Am. Chem. Soc. 2012, 134, 2750.
(16) Puglisi, A.; Annunziata, R.; Benaglia, M.; Cozzi, F.; Gervasini, A.;
Bertacche, V.; Sala, M. C. Adv. Synth. Catal. 2009, 351, 219.
(17) See ref 10 and also: Puglisi, A.; Benaglia, M.; Cinquini, M.;
Cozzi, F.; Celentano, G. Eur. J. Org. Chem. 2004, 567.
(18) Ahrendt, K. A.; Borths, C. J.; MacMillan, D. W. C. J. Am. Chem.
Soc. 2000, 122, 4243.
(19) Asefa, T.; MacLachlan, M. J.; Coombs, N.; Ozin, G. A. Nature
1999, 402, 867.
(20) Athens, G. L.; Shayib, R. M.; Chmelka, B. F. Curr. Opin. Colloid
Interface Sci. 2009, 14, 281.
(21) (a) Huh, S.; Wiench, J. W.; Ji-Chul, Y.; Pruski, M.; Lin, V. S.-Y.
Chem. Mater. 2003, 15, 4247. (b) Huh, S.; Chen, H.-T.; Wiench, J. W.;
Pruski, M.; Lin, V. S.-Y. Angew. Chem., Int. Ed. 2005, 44, 1826.
(22) If a smaller amount of catalyst is used (10 mol %), the reaction
is very slow and a longer reaction time is needed to reach a high
conversion.
(23) A different procedure for recycling the catalyst was then
attempted: after the Diels−Alder reaction, catalyst 3 was recovered by
centrifugation, suspended in DMF, and stirred at 80 °C for 6 h with
the aim of dissolving the insoluble organic residues (mainly
cyclopentadiene dimer), which could occlude the mesopores. After
this treatment and after addition of HBF4, catalyst 3 was reused in the
cycloaddition of cyclopentadiene and cinnamic aldehyde, affording the
products in 30% yield.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail for A.P.: alessandra.puglisi@unimi.it.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This paper is dedicated to Professor Maurizio Prato on the occasion
of his C60th birthday. This work was supported by the MIUR-
PRIN (Nuovi metodi catalitici stereoselettivi e sintesi stereoselettiva
di molecole funzionali), Cariplo Foundation (Milano) within the
project 2011-0293 “Novel chiral recyclable catalysts for one-pot,
multi-step synthesis of structurally complex molecules”, and FIRB
project RBFR10BF5 V “Multifunctional hybrid materials for the
development of sustainable catalytic processes”. M.B. thanks the
COST action CM9505 “ORCA” Organocatalysis.
REFERENCES
■
(1) (a) Nefzi, A.; Ostresh, J. M.; Houghten, R. A. Chem. Rev. 1997,
97, 449. (b) Plante, O. J.; Palmacci, E. R.; Seeberger, P. H. Science
2001, 291, 1523. (c) Myers, R. M.; Roper, K. A.; Baxendale, I. R.; Ley,
S. V. The evolution of immobilised reagents and their application in
flow chemistry for the synthesis of natural products and
pharmaceutical compounds. In Modern Tools for the Synthesis of
Complex Bioactive Molecules; Cossy, J., Arseniyadis, S., Eds.; Wiley:
New York, 2012; pp 359−394. (d) Baxendale, I. R.; Ley, S. V.
Heterogeneous Reactions. In New Avenues to Efficient Chemical
Synthesis, Emerging Technologies; Seeberger, P. H., Blume, T., Eds.;
Springer-Verlag: Berlin/Heidelberg, 2007; pp 151−185. (e) Solid-
Phase Organic Synthesis: Concepts, Strategies, and Applications; Toy, P.
H., Lam, Y., Eds.; Wiley: New York, 2012.
(24) In order to overcome the recycling problems, silica-supported
imidazolidinones have found application in continuous flow chemistry
processes: Chiroli, V.; Benaglia, M.; Cozzi, F.; Puglisi, A.; Annunziata,
R.; Celentano, G. Org. Lett. 2013, 15, 3590.
(25) Shi, J. Y.; Wang, C. A.; Li, Z. J.; Wang, Q.; Zhang, Y.; Wang, W.
Chem. Eur. J. 2011, 17, 6206.
NOTE ADDED AFTER ASAP PUBLICATION
In the version reposted on November 15, 2013 there are text
changes and updates to references 14 and 25.
■
(2) Lee, A.; Breitenbucher, J. G. Curr. Opin. Drug Discovery Dev.
2003, 6, 494.
(3) Jas, G.; Kirschning, A. Chem. Eur. J. 2003, 9, 5708.
(4) Solinas, A.; Taddei, M. Synthesis 2007, 2409.
(5) Kirschning, A.; Solodenko, W.; Mennecke, K. Chem. Eur. J. 2006,
12, 5972.
(6) Trindade, A. F.; Afonso, C. A. M. Chem. Rev. 2009, 109, 3401.
(7) (a) Recoverable and Recyclable Catalysts; Benaglia, M., Ed.; Wiley:
Chichester, U.K., 2009. (b) Haraguchi, N.; Itsuno, S. Polymeric Chiral
Catalyst Design and Chiral Polymer Synthesis; Wiley: Chichester, U.K.,
2011. (c) Handbook of Asymmetric Heterogeneous Catalysts; Ding, K. J.,
Uozomi, F. J. K., Eds.; Wiley-VCH: Weinheim, Germany, 2008.
(8) Reviews on supported organic catalysts: (a) Benaglia, M.; Puglisi,
A.; Cozzi, F. Chem. Rev. 2003, 103, 3401. (b) Cozzi, F. Adv. Synth
Catal. 2006, 348, 1367. (c) Benaglia, M. New J. Chem. 2006, 30, 1525.
(9) Lelais, G.; MacMillan, D. W. C. Aldrichim. Acta 2006, 39, 79.
(10) (a) Benaglia, M.; Celentano, G.; Cinquini, M.; Puglisi, A.; Cozzi,
F. Adv. Synth. Catal. 2002, 344, 149. (b) Guizzetti, S.; Benaglia, M.;
Siegel, J. S. Chem. Commun. 2012, 48, 3188.
(11) Selkala, S. A.; Tois, J.; Pihko, P. M.; Koskinen, A. M. P. Adv.
̈
̈
Synth. Catal. 2002, 344, 941.
(12) Zhang, Y.; Zhao, L.; Lee, S. S.; Ying, J. Y. Adv. Synth. Catal.
2006, 348, 2027.
(13) Hagiwara, H.; Kuroda, T.; Hoshi, T.; Suzuki, T. Adv. Synth.
Catal. 2010, 352, 909.
11334
dx.doi.org/10.1021/jo401852v | J. Org. Chem. 2013, 78, 11326−11334