Acid-base bifunctional and dielectric outer-sphere effects in heterogeneous catalysis: A comparative investigation of model primary amine catalysts

Article abstract of DOI:10.1021/ja057395c

Dielectric and acid-base bifunctional effects are elucidated in heterogeneous aminocatalysis using a synthetic strategy based on bulk silica imprinting. Acid-base cooperativity between silanols and amines yields a bifunctional catalyst for the Henry reaction that forms α,β- unsaturated product via quasi-equilibrated iminium intermediate. Solid-state UV/vis spectroscopy of catalyst materials treated with salicylaldehyde demonstrates zwitterionic iminium ion to be the thermodynamically preferred product in the bifunctional catalyst. This product is observed to a much lesser extent relative to its neutral imine tautomer in primary amine catalysts having outer-sphere silanols partially replaced by aprotic functional groups. One of these primary amine catalysts, consisting of a polar outer-sphere environment derived from cyano-terminated capping groups, has activity comparable to that of the bifunctional catalyst in the Henry reaction, but instead forms the β-nitro alcohol product in high selectivity (～99%). This appears to be the first observation of selective alcohol formation in primary amine catalysis of the Henry reaction. A primary amine catalyst with a methyl-terminated outer-sphere also produces alcohol, albeit at a rate that is 50-fold slower than the cyano-terminated catalyst, demonstrating that outer-sphere dielectric constant affects catalyst activity. We further investigate the importance of organizational effects in enabling acid-base cooperativity within the context of bifunctional catalysis, and the unique role of the solid surface as a macroscopic ligand to impose this cooperativity. Our results unequivocally demonstrate that reaction mechanism and product selectivity in heterogeneous aminocatalysis are critically dependent on the outer-sphere environment.

Full text of DOI:10.1021/ja057395c

Published on Web 02/24/2006
Acid-Base Bifunctional and Dielectric Outer-Sphere Effects in
Heterogeneous Catalysis: A Comparative Investigation of
Model Primary Amine Catalysts
John D. Bass, Andrew Solovyov, Andrew J. Pascall, and Alexander Katz*
Contribution from the Department of Chemical Engineering, UniVersity of California at
Berkeley, Berkeley, California 94720-1462
Received October 29, 2005; E-mail: katz@cchem.berkeley.edu
Abstract: Dielectric and acid-base bifunctional effects are elucidated in heterogeneous aminocatalysis
using a synthetic strategy based on bulk silica imprinting. Acid-base cooperativity between silanols and
amines yields a bifunctional catalyst for the Henry reaction that forms R,â-unsaturated product via quasi-
equilibrated iminium intermediate. Solid-state UV/vis spectroscopy of catalyst materials treated with
salicylaldehyde demonstrates zwitterionic iminium ion to be the thermodynamically preferred product in
the bifunctional catalyst. This product is observed to a much lesser extent relative to its neutral imine tautomer
in primary amine catalysts having outer-sphere silanols partially replaced by aprotic functional groups.
One of these primary amine catalysts, consisting of a polar outer-sphere environment derived from cyano-
terminated capping groups, has activity comparable to that of the bifunctional catalyst in the Henry reaction,
but instead forms the â-nitro alcohol product in high selectivity (∼99%). This appears to be the first
observation of selective alcohol formation in primary amine catalysis of the Henry reaction. A primary amine
catalyst with a methyl-terminated outer-sphere also produces alcohol, albeit at a rate that is 50-fold slower
than the cyano-terminated catalyst, demonstrating that outer-sphere dielectric constant affects catalyst
activity. We further investigate the importance of organizational effects in enabling acid-base cooperativity
within the context of bifunctional catalysis, and the unique role of the solid surface as a macroscopic ligand
to impose this cooperativity. Our results unequivocally demonstrate that reaction mechanism and product
selectivity in heterogeneous aminocatalysis are critically dependent on the outer-sphere environment.
Introduction
bifunctional catalysts can strongly depend on the separation
distance between acid and base groups.8,9,16-21 This was
Acid-base cooperativity is commonly invoked in enzymatic
and catalytic antibody catalysis.1-7 This paradigm of biological
catalysts has inspired the synthesis of homogeneous bifunctional
catalysts with superior activity and selectivity, including enan-
tioselectivity, relative to their individual acid and base
counterparts.8-16 However, the function of these homogeneous
illustrated by Swain and Brown, who showed that while
2-hydroxypyridine catalyzes the mutarotation of tetramethyl-
glucose via an acid-base bifunctional mechanism, 3- and
4-hydroxypyridines are 1000-fold less active and exhibit kinetics
consistent with general catalysis.8 Recent work in the synthesis
of homogeneous bifunctional catalysts demonstrates a pro-
nounced distance sensitivity for acid-base cooperativity.8,9,18-21
A heterogeneous organic-inorganic bifunctional catalyst
could, in principle, overcome this sensitivity by providing a near-
continuous range of acid-base distances.19 This could increase
bifunctional catalyst versatility by providing various optimal
distances for cooperativity, as illustrated in Scheme 1. Such a
catalyst would be particularly advantageous for multistep
reactions where several kinetically relevant steps have unique
distance requirements for acid-base coopertivity. However, to
the best of our knowledge, there are no examples of hybrid
organic-inorganic materials rigorously proven to act as acid-
(1) Mosier, N. S.; Hall, P.; Ladisch, C. M.; Ladisch, M. R. AdV. Biochem.
Eng./Biotechnol. 1999, 65, 23-40.
(2) Mccarter, J. D.; Withers, S. G. Curr. Opin. Struct. Biol. 1994, 4, 885-
892.
(3) Debler, E. W.; Ito, S.; Seebeck, F. P.; Heine, A.; Hilvert, D.; Wilson, I. A.
Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 4984-4989.
(4) Hoffmann, T.; Zhong, G. F.; List, B.; Shabat, D.; Anderson, J.; Gramatikova,
S.; Lerner, R. A.; Barbas, C. F. J. Am. Chem. Soc. 1998, 120, 2768-2779.
(5) Hennig, M.; Darimont, B. D.; Jansonius, J. N.; Kirschner, K. J. Mol. Biol.
2002, 319, 757-766.
(6) Kartha, G.; Bello, J.; Harker, D. Nature 1967, 213, 862-865.
(7) Carter, P.; Wells, J. A. Nature 1988, 332, 564-568.
(8) Swain, C. G.; Brown, J. F. J. Am. Chem. Soc. 1952, 74, 2534-2537.
(9) Okino, T.; Hoashi, Y.; Furukawa, T.; Xu, X. N.; Takemoto, Y. J. Am.
Chem. Soc. 2005, 127, 119-125.
(10) Okino, T.; Hoashi, Y.; Takemoto, Y. J. Am. Chem. Soc. 2003, 125, 12672-
12673.
(11) Sakthivel, K.; Notz, W.; Bui, T.; Barbas, C. F. J. Am. Chem. Soc. 2001,
123, 5260-5267.
(17) Breslow, R.; Graff, A. J. Am. Chem. Soc. 1993, 115, 10988-10989.
(18) Hine, J.; Li, W. S. J. Org. Chem. 1975, 40, 2622-2626.
(19) Hine, J.; Cholod, M. S.; King, R. A. J. Am. Chem. Soc. 1974, 96, 835-
845.
(12) List, B. Synlett 2001, 1675-1686.
(13) Wang, W.; Wang, J.; Li, H. Angew. Chem., Int. Ed. 2005, 44, 1369-1371.
(14) Saito, S.; Yamamoto, H. Acc. Chem. Res. 2004, 37, 570-579.
(15) Hine, J. Acc. Chem. Res. 1978, 11, 1-7.
(20) Calter, M. A.; Orr, R. K. Tetrahedron Lett. 2003, 44, 5699-5701.
(16) Anslyn, E.; Breslow, R. J. Am. Chem. Soc. 1989, 111, 8931-8932.
(21) Tanabe, K.; Yamaguchi, T. Catal. Today 1994, 20, 185-198.
9
10.1021/ja057395c CCC: $33.50 © 2006 American Chemical Society
J. AM. CHEM. SOC. 2006, 128, 3737-3747
3737

A R T I C L E S
Bass et al.
Scheme 1. The Silica Surface as a Macroscopic Ligand Providing
a Range of Acid-Base Distances for Bifunctional Cooperativity
chlorosilane at identical surface coverage yields the polar/
nonacidic material 3. Because these latter two materials are
derived from the same parent mesoporous material 1, the
structure of the silica network and the relative positioning of
active sites are identical. Carbamates in materials 1, 2, and 3
are subsequently thermally deprotected by mild heating to form
catalytic materials 4, 5, and 6 (Chart 1).34
Carbamate-deprotection kinetics provides the first measure
of the importance of outer-sphere acid-base cooperativity and
dielectric environment on reactivity at the active site. We also
use salicylaldehyde as a probe molecule in binding experiments
aimed at characterizing outer-sphere acidity and dielectric
environment of primary amines in 4, 5, and 6. In addition,
materials 4, 5, and 6 are used as catalysts for the Michael, Henry,
and Knoevenagel reactions. Results are compared to relevant
homogeneous models 7 and 8, which consist of a primary amine
tethered to a phenolic oxygen on the lower rim of a calix[4]arene
scaffold. Such calix[4]arene triols were shown to be Bronsted
acid centers with a pKa of 6.9;35 thus, they are a good model
for tethered amines on silica, because silanols on silica surfaces
have an average pKa value of 7.1.36,37 Related calix[4]arene triols
show acid-base cooperativity in aldol catalysis,38 with the
length of the amine tether strongly influencing activity.38 We
compare 7 and 8 with catalyst 4 to highlight the role of the
heterogeneous surface as an organizational scaffold that enables
acid-base bifunctional cooperativity in catalysis. Our results
provide new mechanistic insights into primary amine reactivity
as controlled by the outer-sphere environment, which are
significant given the prevalence of primary amine catalysis,
including in biological systems where lysine residues play a
central catalytic role.39,40
base bifunctional catalysts. This is due to a materials limitation
caused by the inability to selectively remove the acid or base
component of the active site without concurrently modifying
the catalyst structure or the relative positioning of active
sites.21-24 Both bulk structural characteristics25 and cooperative
effects between proximal functional groups26 can strongly
influence catalyst activity. Also, changes to the outer-sphere
dielectric environment must be explicitly addressed when
selectively removing acid or base components via a materials
synthesis strategy. These dielectric effects play an important
role in catalysis.27-33 In biological systems, for example, the
outer-sphere dielectric environment created by the hydrophobic
protein interior can change the reactivity of primary amines by
lowering their pKa, leading to highly nucleophilic amine
groups.30-33
Our goal is to assess how the outer-sphere dielectric environ-
ment and acid-base cooperativity independently contribute to
chemical reactivity. Materials syntheses based on bulk silica
imprinting allow us to systematically vary the composition of
the outer sphere while keeping other material properties constant.
Our approach involves copolymerization of a carbamate imprint
with tetraethyl orthosilicate to form hybrid organic-inorganic
sol-gel material 1. Material 1 consists of a silanol-rich polar/
acidic mesoporous network that is the direct result of the sol-
gel synthesis. It is subsequently modified via capping of silanols
to tailor the outer-sphere environment. The isolated active sites
remain chemically protected as carbamates during synthesis and
capping; thus, changes to the outer-sphere environment do not
affect the composition of the ultimate catalytically active primary
amine. The nonpolar/nonacidic material, 2, is synthesized by
capping accessible silanols with excess dimethyl butyl chlo-
rosilane. Capping with the size-similar dimethyl cyanopropyl
Results and Discussion
Thermolytic carbamate decomposition in materials 1, 2, and
3 proceeds via a charge-separated transition state involving
protonation of the carbonyl oxygen.41 Based on this mechanism,
both the dielectric environment and the presence of an acidic
silanol functionality in the outer sphere should affect thermolysis
kinetics. Figure 1 shows the derivative of the weight loss during
carbamate thermolysis. The maximum rate of thermolysis for
1, which contains acidic silanols, occurs at 198 °C, about 46
°C lower than for nonacidic materials 2 and 3. A smaller
(22) Iglesia, E.; Barton, D. G.; Biscardi, J. A.; Gines, M. J. L.; Soled, S. L.
Catal. Today 1997, 38, 339-360.
(23) Tanabe, K. J. Chin. Chem. Soc. 1998, 45, 597-602.
(24) Yamaguch, T.; Sasaki, H.; Tanabe, K. Chem. Lett. 1973, 1017-1018.
(25) Macquarrie, D. J.; Maggi, R.; Mazzacani, A.; Sartori, G.; Sartorio, R. Appl.
Catal., A 2003, 246, 183-188.
(26) Dufaud, V.; Davis, M. E. J. Am. Chem. Soc. 2003, 125, 9403-9413.
(27) Thorn, S. N.; Daniels, R. G.; Auditor, M. T. M.; Hilvert, D. Nature 1995,
373, 228-230.
(28) Kemp, D. S.; Cox, D. D.; Paul, K. G. J. Am. Chem. Soc. 1975, 97, 7312-
7318.
(34) Bass, J. D.; Katz, A. Chem. Mater. 2003, 15, 2757-2763.
(35) Araki, K.; Iwamoto, K.; Shinkai, S.; Matsuda, T. Bull. Chem. Soc. Jpn.
1990, 63, 3480-3485.
(29) Harris, T. K. T., G. J. Life 2002, 53, 85-98.
(30) Urry, D. W. P., S. Q.; Parker, T. M.; Gowda, D. C.; Harris, R. D. Angew.
Chem., Int. Ed. Engl. 1993, 32, 1440-1442.
(36) Hair, M. L.; Hertl, W. J. Phys. Chem. 1970, 74, 91-94.
(37) Marshall, K.; Ridgewel, G.; Rocheste, C.; Simpson, J. Chem. Ind. 1974,
775-776.
(31) Urry, D. W. G., D. C.; Peng, S. Q.; Parker, T. M.; Harris, R. D. J. Am.
Chem. Soc. 1992, 114, 8716-8717.
(32) Karlstrom, A. Z., G.; Rader, C.; Larsen, N. A.; Heine, A.; Fuller, R.; List,
B.; Tanaka, F.; Wilson, I. A.; Barbas, C. F.; Lerner, R. A. Proc. Natl.
Acad. Sci. U.S.A. 2000, 97, 3878-3883.
(38) Liu, L.; Yang, Z. X.; Wang, Y. L.; Chen, S. H. Chin. Chem. Lett. 2000,
11, 485-488.
(39) Jencks, W. P. Annu. ReV. Biochem. 1963, 32, 639-676.
(40) Speck, J. C.; Rowley, P. T.; Horecker, B. L. J. Am. Chem. Soc. 1963, 85,
1012-1013.
(33) Barbas, C. F. H., A.; Zhong, G.; Hoffmann, T.; Gramatikova, S.;
Bjornestedt, R.; List, B.; Anderson, J.; Stura, E. A.; Wilson, I. A.; Lerner,
R. A. Science 1997, 278, 2085-2092.
(41) Ashcroft, S. J.; Thorne, M. P. Can. J. Chem. 1972, 50, 3478-3487.
9
3738 J. AM. CHEM. SOC. VOL. 128, NO. 11, 2006

Outer-Sphere Effects in Heterogeneous Catalysis
A R T I C L E S
Chart 1
difference caused by dielectric effects is apparent in the slightly
lower thermolysis temperature of 3 versus 2. Importantly, only
a homogeneous population of sites is observed for each material.
This is observed by a single event in the TGA trace during
deprotection (the deviation from a first-order trace for 1 is
observed due to minor weight loss from desorption of phys-
isorbed water). This indicates that the majority of active sites
within each material experience the same outer-sphere environ-
ment. Such a uniform site distribution suggests single-site
behavior and allows us to draw conclusions about the entire
population of active sites in the reactivity studies below.
Activation energies for carbamate thermolysis in materials
1, 2, and 3 were calculated from TGA data using the method
of Redhead42 (Table 1) at conditions where the rate of reaction
was limited solely by chemical kinetics.43 Assuming a value
for the frequency factor of 1013 s-1, the activation energies were
measured to be 26.7 ( 0.1, 29.9 ( 0.15, and 29.4 ( 0.15 kcal/
mol for 1, 2, and 3, respectively. The slightly lower activation
energy for carbamate thermolysis in the polar/nonacidic 3 as
compared to nonpolar/nonacidic 2 is attributed to the higher
dielectric environment in the outer-sphere capable of stabilizing
charge separation in the transition state. Acidity has a larger
effect, decreasing the activation barrier for carbamate thermoly-
sis in 1 by 2.7 kcal/mol. For this to occur, acidic silanols must
be present in the outer sphere in a position to interact with the
carbonyl group during thermolysis (Scheme 2). The lower
activity for thermolysis of 2 and 3 relative to 1 is not due to a
general lack of silanols in these materials. Total organic content
via combustion analysis shows that capping consumes 1.1
mmol/g of silanols on the surface. This represents 30% of the
silanols shown to be in 1 by infrared spectroscopy (Supporting
Information). Thus, most silanols remain unreacted during
capping of 1 in synthesis of 2 and 3, and the total number of
silanols remains in greater than 10-fold excess relative to
carbamate, with remaining silanols less active for facilitating
thermolysis, as reflected by the greater activation energy for
materials 2 and 3.
(42) Redhead, P. A. Vacuum 1962, 12, 203-211.
(43) No variations were observed using heating rates of 0.5-5 °C/min. Larger
particles (30-150 µm) were compared to smaller (<30 µm) and mixed
batches and were found to give equal values.
We used the site-directed probe salicylaldehyde to investigate
the acidity and dielectric effects of the outer-sphere environment
9
J. AM. CHEM. SOC. VOL. 128, NO. 11, 2006 3739

A R T I C L E S
Bass et al.
Figure 2. Solid-state diffuse-reflectance UV/visible spectra of (A) polar/
acidic catalyst 4, (B) nonpolar/nonacidic catalyst 5, and (C) polar/nonacidic
catalyst 6 after reaction with salicylaldehyde to form a solvochromatic
indicator. The indicator in 4 shows a pronounced zwitterion band near 400
nm due to the acidic silanols. The upper two spectra show a strong band
from the phenolic form at 315 nm of the indicator. Data are represented in
Kubelka Munk units.
Figure 1. Thermal gravimetric analysis of carbamate deprotection in
materials 1, 2, and 3. The temperature of the maximum thermolysis rate is
significantly lower in the polar/acidic 1 (A) at 198 °C as compared to the
polar/nonacidic 3 (B) at 244 °C and the nonpolar/nonacidic material 2 (C)
at 251 °C. N2 flow of 20 mL min-1 with a programmed heating rate of 1
°C per min.
Scheme 3. The Reaction between Salicylaldehyde and a Primary
Amine Can Form (A) the Zwitterionic Iminium Tautomer in the
Presence of Acidic Silanols as in 4 or (B) the Neutral Imine in the
Absence of Acidic Silanols as in 5 and 6
Table 1. Kinetics of Immobilized Carbamate Deprotection via
Thermolysis
temperature at
maximum rate of
thermolysisa
material
activation energyb
1
2
3
198 °C
251 °C
244 °C
26.7 ( 0.1 kcal/mol
29.9 ( 0.15 kcal/mol
29.4 ( 0.15 kcal/mol
a Heating rate of 1 °C per min. b Based on the method of Redhead,42
frequency factor assumed as 1013 s-1; error analysis on the basis of the
variation in the measured data.
Scheme 2. Thermolytic Deprotection of the Carbamate Protecting
Group Is Facilitated in the Presence of Acidic Silanolsa
zwitterionic iminium tautomer in Scheme 3. Note that by using
a slightly different materials synthesis recipe for sol-gel
hydrolysis and condensation, we have previously prepared a
material similar in composition to 4 that showed even less
neutral phenolic relative to zwitterionic contribution in Figure
2A.44 Thus, this product ratio can vary depending on subtle
differences in the materials synthesis procedure, further illustrat-
ing the importance of deriving all materials from the same parent
material. (For this reason, our procedures are reported in detail
in the Supporting Information.)
These results are consistent with previous studies with imines
derived from salicylaldehyde in homogeneous media. The
zwitterionic iminium tautomer was observed on dissolution in
acidic protic solvents, such as highly fluorinated alcohols, and
in aprotic solvents the neutral phenolic product was favored.45-47
For salicylideneaniline, the stability of the zwitterionic species
has been correlated to the Kamlet-Taft solvent parameter R,
which relates to the H-bond donating ability of the solvent.47
a The silanols must have the correct geometrical orientation relative to
the carbamate functional groups to facilitate the reaction.
surrounding primary amines in materials 4, 5, and 6, as well as
in homogeneous models 7 and 8. The condensation of primary
amines with salicylaldehyde can form two tautomeric prod-
ucts: a neutral phenolic imine with an absorption band at 310-
350 nm and a zwitterionic tautomer that absorbs at 400-450
nm.34,44 Figure 2 shows the diffuse reflectance UV/visible
spectra of 4, 5, and 6 after contact with salicylaldehyde. The
strong band at 400 nm represents the zwitterionic iminium
tautomer resulting from the outer-sphere acidity in 4. In 5 and
6, this band is much weaker than that at 315 nm representing
the neutral phenolic form of the imine. Acidic silanols in
material 4 must be present within the outer sphere of the active
site so as to hydrogen bond to the phenolic oxygen to form
(45) Casades, I.; Alvaro, M.; Garcia, H.; Pillai, M. N. Eur. J. Org. Chem. 2002,
2074-2079.
(46) Turbeville, W.; Dutta, P. K. J. Phys. Chem. 1990, 94, 4060-4066.
(47) Dutta, P. K.; Turbeville, W. J. Phys. Chem. 1991, 95, 4087-4092.
(44) Bass, J. D.; Anderson, S. L.; Katz, A. Angew. Chem., Int. Ed. 2003, 42,
5219-5222.
9
3740 J. AM. CHEM. SOC. VOL. 128, NO. 11, 2006

Outer-Sphere Effects in Heterogeneous Catalysis
A R T I C L E S
O(3). The calixarenes, in the cone conformation, are held
together in pairs by strong hydrogen bonds between the primary
ammonium cations and the phenolate anions on the adjacent
calixarene. The distance from N(1) to O(3) on the adjacent
calixarene in the dimer is 2.65(4) Å, and one of the ammonium
protons lies near the N-O vector. In a pattern differing markedly
from the typical pattern observed in other monosubstituted
calix[4]arenes without base substituents,35 two intramolecular
hydrogen bonds from O(2) and O(4) to O(3) are observed. The
O-O distance is 2.58(5) Å for each, and the located hydrogen
atoms lie along the corresponding vectors. There is also a
moderately strong bifurcated intramolecular hydrogen bond from
the ammonium N(1) to the phenol oxygen O(2) (N-O distance
) 2.90(3) Å) and to O(1) (N-O distance ) 2.79(5) Å). These
results show an acidic, hydrogen-bonded phenol triol that has
protonated the primary amine, allowing the formation of strong
intermolecular and intramolecular hydrogen-bonding interac-
tions.
Evidence for acid-base cooperativity in the homogeneous
models 7 and 8 upon reaction with salicylaldehyde was
investigated. Phenolic imines 9 and 10 were isolated upon
treatment of 7 and 8 with salicylaldehyde. These imines, despite
retaining the acidic phenolic triol, show no zwitterionic tautomer
in aprotic solvents (Figure 4). Only upon solvation in a highly
protic environment does the zwitterion tautomer become favored
as a result of intermolecular solvent interactions. Comparing
these results with the heterogeneous acid-base bifunctional
material 4 treated with salicylaldehyde confirms the importance
of the silanol-rich silica surface as a macroscopic acidic ligand
that presents the primary amines with a continuous acidic surface
for facilitating acid-base cooperativity.
Figure 3. Structure of the homogeneous acid-base analogue 7 showing
(a) the organized dimer and (b) the hydrogen-bonding pattern within a
monomer unit comprising the dimer in (a). The dimer is held together by
a strong intermolecular interaction between the primary ammonium cation
and the phenolate anion on oxygen O(3) of the calixarene lower rim.
No correlation was previously observed for the parameter Π*,
which relates to the ability of a solvent to stabilize charge via
dielectric effects,47 which is consistent with the same observed
preference for the neutral phenolic form in materials 5 and 6.
The spectra in Figure 2 demonstrate that the dielectric
environment in the nonpolar/nonacidic 5 is lower than in polar
materials 4 and 6. The wavelength maximum of the zwitterionic
band of the bound chromophore undergoes a hypsochromic shift
of 6 nm from the nonpolar 5 (402 nm) to the polar materials 4
and 6 (both at 396 nm). This shift is typical of π-π* transitions
in zwitterionic species, because the excited state has a smaller
dipole moment than the ground state.48 This decreases the energy
required for excitation in nonpolar 5 relative to the polar
materials 4 and 6.
The acid-base bifunctional nature of the homogeneous
models 7 and 8 can be hypothesized on the basis of the known
acidity of lower rim calix[4]arene triols16 and the interactions
of such acidic OH nests with primary amines.49-51 The acid-
base bifunctional nature is also observed in the single-crystal
X-ray diffraction structure of 7 (Figure 3). During structure
refinement, the ammonium and hydroxyl hydrogens atoms were
located in a difference Fourier map and included in their
observed positions normalized to standard N-H and O-H
distances. No peaks were observed near the phenolate oxygen
We next examine probe reactions with turnovers to investigate
outer-sphere effects on primary amine catalysis. The Michael
addition of malononitrile to â-nitrostyrene is catalyzed by
general bases via an ion-pair mechanism (Scheme 4), where
the function of the general base is to deprotonate the malono-
nitrile to form carbanion intermediates.52 The results of this
reaction under anhydrous conditions are shown in Figure 5. The
higher activity of 6 over 5 (a factor of 7) is consistent with the
ability of a high-dielectric outer-sphere environment to facilitate
ion-pair formation and charge separation in the transition state.
Acid-base bifunctional effects were examined independently
from outer-sphere dielectric contributions on catalysis using the
Henry reaction as a probe. We hypothesize that the Henry
reaction may potentially proceed via two mechanisms: an imine
mechanism (Scheme 5) and an ion-pair mechanism (Scheme
(48) Reichardt, C. Chem. ReV. 1994, 94, 2319-2358.
(49) El Nahhal, I. M.; Chehimi, M. M.; Cordier, C.; Dodin, G. J. Non-Cryst.
Solids 2000, 275, 142-146.
(50) Yang, J. J.; ElNahhal, I. M.; Chuang, I. S.; Maciel, G. E. J. Non-Cryst.
Solids 1997, 209, 19-39.
(52) Carey, F. A.; Sundberg, R. J. AdVanced Organic Chemistry, Part B, 3rd
(51) Horr, T. J.; Arora, P. S. Colloids Surf., A 1997, 126, 113-121.
ed.; Plenum Press: New York, 1990.
9
J. AM. CHEM. SOC. VOL. 128, NO. 11, 2006 3741

A R T I C L E S
Bass et al.
Scheme 4. Ion-Pair Mechanism of Catalysis by Amine-Functionalized Silica for the Michael Reaction
6). The mechanism involving covalent imine intermediates is
reported to account for primary amine catalysis in homo-
geneous53-55 and enzymatic39,40 systems. In this mechanism,
both imine formation15,56,57 and imine protonation to form
iminum ions55,58 are promoted by acid-base bifunctionality.
Formation of an iminum ion, as in the salicylaldehyde study
above, creates a carbon center that is significantly more
electrophilic than the carbonyl compound.52,55,58 This facilitates
attack by the carbon nucleophile during Henry catalysis.
In the presence of bases for which imine formation is
kinetically unfavorable, we postulate that the Henry reaction
may produce the â-nitro alcohol product via an ion-pair
mechanism. This mechanism has been proposed previously59-61
for primary amine catalysis on silica, but it has remained
unproven. Crucially, because the olefin and alcohol products
do not interconvert under reaction conditions,62 the product
distribution directly reflects the reaction mechanism: the R,â-
unsaturated product forms via the imine mechanism (Scheme
5) and the â-nitro alcohol via the ion-pair mechanism (Scheme
6).
The Henry reaction of nitromethane and 4-nitrobenzaldehyde
on catalysts 4, 5, and 6 shows a marked, and unprecedented,
reversal in selectivity between the polar/acidic catalyst 4 and
the polar/nonacidic catalyst 6. These kinetic data, measured
using 1H NMR spectroscopy, are shown in Figure 6. In catalyst
4, acid-base bifunctional cooperativity among silanols and the
primary amines results in a lower barrier for imine formation
and the selective formation of product 11. This product is formed
at a rate that is 30-fold greater than for the nonacidic catalysts
5 and 6. The observed zero order kinetic dependence on
aldehyde reactant concentration up to nearly 90% conversion
of aldehyde indicates that imine formation is quasi-equilibrated
and energetically favored during catalysis, leading to saturation
of active sites for catalyst 4.
Figure 4. UV/visible spectra of the homogeneous acid-base analogue 9
in three different solvents showing absorbance (au) as a function of
wavelength. In acetonitrile (A), the imine exists in the phenolic form, despite
being pendent to an acidic species. The zwitterionic tautomer is only formed
in strongly hydrogen bond donating solvents, being fractionally present in
methanol (B), and predominant in pentafluoro-1-propanol (C). Similar results
are obtained for the calixarene 10 that contains a four-carbon tether. The
concentration is 60 µM.
In stark contrast with the acid-base bifunctional imine
mechanism on catalyst 4, catalyst 6 favors instead the ion-pair
mechanism and forms the â-nitro alcohol 12 with 99% selectiv-
(53) Westheimer, F. H.; Cohen, H. J. Am. Chem. Soc. 1938, 60, 90-94.
(54) Spencer, T. A.; Neel, H. S.; Flechtne, T.; Zayle, R. A. Tetrahedron Lett.
1965, 3889-3897.
(55) Crowell, T. I.; Peck, D. W. J. Am. Chem. Soc. 1953, 75, 1075-1077.
(56) Hine, J.; Cholod, M. S.; Chess, W. K. J. Am. Chem. Soc. 1973, 95, 4270-
4276.
(57) Hine, J.; Via, F. A. J. Org. Chem. 1977, 42, 1972-1978.
(58) Santerre, G. M.; Hansrote, C. J.; Crowell, T. I. J. Am. Chem. Soc. 1958,
80, 1254-1257.
(59) Angeletti, E.; Canepa, C.; Martinetti, G.; Venturello, P. J. Chem. Soc., Perkin
Trans. 1 1989, 105-107.
(60) Choudary, B. M.; Kantam, M. L.; Reddy, C. V.; Rao, K. K.; Figueras, F.
Green Chem. 1999, 1, 187-189.
(61) Huh, S.; Chen, H. T.; Wiench, J. W.; Pruski, M.; Lin, V. S. Y. J. Am.
Chem. Soc. 2004, 126, 1010-1011.
Figure 5. Conversion of trans-â-nitrostyrene for its base-catalyzed Michael
addition with malononitrile. The high dielectric catalyst 6 (2) exhibits a
7-fold higher activity as compared to the nonpolar catalyst 5 (9). This rate
enhancement is due to the changes in the outer-sphere dielectric environment
surrounding the amine active site. Data for catalyst 4 (O) are also shown
as well as solid lines fitting to a first-order kinetics model for 4 and 5. The
yield of the 2-cyano-4-nitro-3-phenylbutyronitrile product at 100% nitrosty-
(62) Interconversion was ruled out by conducting the following experiments
under dry conditions. (1) Reaction in the presence of 6 and subsequent
filtration and transfer via syringe to 4 after 60% conversion of aldehyde
(54% yield of 12). No decrease in 12 was observed after 4 h exposure to
4 and aldehyde conversion to 90%. (2) The reverse catalyst sequence using
identical methods with filtration and transfer at 20% aldehyde conversion
(13% yield of 11). No decrease in 11 was observed after 5.5 h exposure to
6 and aldehyde conversion to 50%.
1
rene conversion via H NMR was 96% ( 3% using catalyst 6.
9
3742 J. AM. CHEM. SOC. VOL. 128, NO. 11, 2006

Outer-Sphere Effects in Heterogeneous Catalysis
A R T I C L E S
Scheme 5. Imine Catalytic Mechanism for the Production of the Dehydrated Product for the Henry Reaction (with Nitromethane) Catalyzed
by Amines on Silicaa
a Both the imine formation step and the attack of the nucleophile on the imine are known to be acid catalyzed.
Scheme 6. Ion-Pair Mechanism of Catalysis by Amine-Functionalized Silica for the Production of â-Nitro Alcohol Product in the Henry
Reaction
ity. As in the Michael reaction, dielectric effects play a
significant role in promoting the ion-pair pathway for the Henry
reaction on nonacidic materials. The rate of formation of â-nitro
alcohol on 6 is ∼50 times greater than that on 5. This again
reflects a high local dielectric environment that stabilizes ion
pair formation and charge separation in the transition state. The
generality of these trends above was investigated by using a
commercial 3-aminopropyl-functionalized material for the same
Henry reaction. This hybrid material consists of a dense
monolayer of 3-aminopropyl groups on silica (1.3 mmol/g). We
have previously shown that the outer-sphere surrounding the
amine active sites in this material behaves as a nonacidic
material for related condensation reactions.44 The absence of
acid sites due the displacement of silanols with the 3-amino-
propyl modifier has been measured using X-ray photoelectron
spectroscopy.49-51 This catalyst forms mainly product 12, but
with lower selectivity to product 12 (85% at 20% conversion)
and at a 4-fold lower rate per amine than catalyst 6 under
identical conditions.
results demonstrate that the reaction mechanism depends
critically on the outer-sphere environment surrounding the
immobilized primary amine. Outer-sphere acidity creates an
active pathway to dehydrated products via imine formation,
while the polar/nonacidic outer-sphere environment of material
6 creates a primary amine that is active and selective for product
formation via ion-pair mechanisms.
Materials 4, 5, and 6 are all active catalysts for Knoevenagel
condensation of aldehydes with activated methylene compounds
(Table 2). Previously, we observed large effects of up to 2 orders
of magnitude difference for the polar/acidic environment found
in catalyst 4 relative to a material that was capped with nonpolar/
nonacidic trimethylsilyl chloride (similar to material 5);44
however, the origin of these effects (dielectric versus acid-
base bifunctional) remained unclear. In this report, we performed
the Knoevenagel condensation of malononitrile with 3-nitroben-
zaldehyde on 4, 5, and 6. Here, the alcohol product is expected
to be unstable under reaction conditions because of the strength
of electron-withdrawing groups in the malononitrile adduct,
leading to the same olefin product in both imine and ion-pair
mechanisms.
Our demonstration that primary amines can function as
general base catalysts for synthesizing products via an ion-pair
mechanism is unique, in lifht of all previous observations of
primary amine catalysis via imine intermediates in both
homogeneous53-55 and enzymatic39,40 systems. A postulated59-61
ion-pair mechanism for amine on silica materials remains to be
demonstrated,63,64 and, to the best of our knowledge, preference
for ion-pair mechanisms over condensation mechanisms is
shown here for the first time for a primary amine catalyst. Our
The results of the Knoevenagel condensation between 3-ni-
trobenzaldehyde and malononitrile on catalysts 4, 5, and 6 are
shown in Figure 7. Reactions were conducted under anhydrous
conditions to avoid kinetic complications caused by water (vide
infra) by adding dehydrated mesoporous silica as a minimally
catalytic desiccant. This also keeps the total number of acidic
silanols in the reaction constant for catalysts 5 and 6 (no silica
was added for catalysis with 4). In this way, all reactions contain
the same active functional groups and differ only in their
(63) Lasperas, M.; Llorett, T.; Chaves, L.; Rodriguez, I.; Cauvel, A.; Brunel,
D. Heterog. Catal. Fine Chem. IV 1997, 108, 75-82.
(64) Demicheli, G.; Maggi, R.; Mazzacani, A.; Righi, P.; Sartori, G.; Bigi, F.
Tetrahedron Lett. 2001, 42, 2401-2403.
9
J. AM. CHEM. SOC. VOL. 128, NO. 11, 2006 3743

A R T I C L E S
Bass et al.
Figure 7. The fractional conversion of 3-nitrobenzaldehyde as a function
of time for the base-catalyzed Knoevenagel condensation with malononitrile.
Outer-sphere differences between the acidic catalyst 4 (O) and the nonacidic
catalysts 5 (9) and 6 (2) lead to significant effects in the observed rate of
reaction. The principal cause of the rate acceleration in 4 results from the
presence of acidic silanols acting in concert with the amine in an acid-
base bifunctional mechanism. The rate difference between polar/nonacidic
6 and the nonpolar/nonacidic 5 shows the secondary effect of the dielectric
environment. Solid lines represent first-order kinetic models with respect
to 3-nitrobenzaldehyde concentration.
Table 2. Summary of Catalysis for Michael, Henry, and
Knoevenagel Reactions under Dry Conditions
Henry product ratiob
a
a
material
Michael kapp
Knoevenagel kapp
[TOF]b
4
5
0.27
0.03
2.9
80:20 [12]
40:60 [0.6]
5 and silicac
6
0.22
Figure 6. Product formation for the Henry reaction between 4-nitroben-
zaldehyde and nitromethane. Outer-sphere differences between the acidic
catalyst 4 (O) and the nonacidic catalysts 5 (9) and 6 (2) lead to significant
effects in the observed selectivity with catalyst 4 producing predominantly
(4:1) R,â-unsaturated product 11 (A) at a rate 30-fold higher than the
nonacidic catalysts. The formation of the â-nitro alcohol product 12 (B) is
sensitive to the dielectric environment, comparing 5 to 6, where the rate
increases by a factor of 50 in the more polar material. The combined product
yield per aldehyde reacted was 95% ( 5% for all catalysts studied. Dashed
lines show the initial TOF.
0.23
0
1:99 [22]
6 and silicac
silicad
0.48
0.13
0:100
[0]
cyano capped silicad
a h-1, using -dCaldehyde/dt ) kappCaldehyde; at low conversion for Michael
reaction. b Ratio of dehydrated to alcohol product (30% aldehyde conver-
sion), TOF: initial number of reaction events h-1 amine site-1 c Mesoporous
.
silica as desiccant; an equivalent number of silanols as 1 was added.
d Materials treated like they have the equivalent number of amines as the
corresponding imprinted catalyst; dehydrated at 250 °C for 4 h prior to
reaction.
organization of these groups. The dehydrated mesoporous silica
itself contributes less than 5% to overall catalytic rate for catalyst
4.
appear in a particle size variation experiment, we synthesized
the nonpolar/nonacidic catalyst 13 using the capping agent
trimethyl chlorosilane. Molecular modeling shows that this
capping group is 40% smaller than the dimethyl butyl chlo-
rosilane used to synthesize the nonpolar/nonacidic catalyst 5.
Weight loss in thermogravimetric analysis shows that materials
5 and 13 have identical capping efficiencies of 1.1 mmol/g (this
is also true for 6). This consistency reflects our ability to
minimize differences between materials via the synthesis
procedures used, which becomes possible because all materials
are derived from the same parent material 1. With an equivalent
fraction of OH groups capped, only the steric bulk at the surface
is different, being significantly smaller in 7 than in 5. Despite
this difference in steric bulk at the surface, the materials behaved
identically for Knoevenagel catalysis (see Supporting Informa-
tion) and have nearly identical activation energies for thermoly-
sis (30.1 ( 0.2 kcal/mol for 13). Also, materials 5 and 6, which
differ in the molecular size of the capping group by less than
5%, showed marked differences in catalytic behavior for all three
model reactions. Taken together, these results demonstrate that
any changes in steric bulk at the catalyst surface caused by
Comparing catalysts 5 and 6 in Figure 7 shows that dielectric
effects increase the rate ∼3-fold. However, the effect of the
acidic silanol groups in 4 is more significant, increasing the
rate by an additional 8-fold over 6 (22-fold faster than 5). The
lower rate in 5 and 6 reflects the removal of the acid
functionality, which, in turn, eliminates acid-base bifunctional
cooperativity. Like in the Henry reaction, the role of the acidic
functional groups in 1 is to divert the reaction through the imine
mechanism via bifunctional catalysis, yielding a higher rate.
These observations again illustrate how acid-base organization
controls bifunctional catalysis, even in heterogeneous systems,
as all reactions contain the same quantity of amines and silanols.
Mass transport effects were ruled out by similar Knoevenagel
reaction rates with large (30-150 µm) and small (<30 µm)
particles.65 Any potential blocking effects of the capping groups
in preventing amine active site accessibility by virtue of their
steric bulk were also ruled out. Because such effects would not
(65) This is the fastest reaction studied in this Article and would be most sensitive
to these effects. These data, along with representative SEM pictures, are
available in the Supporting Information.
9
3744 J. AM. CHEM. SOC. VOL. 128, NO. 11, 2006

Outer-Sphere Effects in Heterogeneous Catalysis
A R T I C L E S
Table 3. Summary of Catalysis for Michael, Henry, and
Knoevenagel Reactions under Ambient Conditions
pronounced for 6, for which the rate decreases by an order of
magnitude for both the Michael and the Henry reactions relative
to anhydrous conditions. Also important is the concomitant
change in selectivity observed in the Henry reaction for material
6 between anhydrous and wet conditions. This decreases from
a greater than 99% selectivity for product 11 under anhydrous
conditions to a selectivity of 40% when small amounts of water
are present. Thus, the decrease in rate of â-nitro alcohol product
formation under wet conditions in 6 is about 50-fold and
indicates that the presence of water specifically inhibits the ion-
pair mechanism. This inhibition is likely to arise from competi-
tive protonation of the amine by water in these materials, as
protonation of amines on silica by water after exposure to humid
air has been observed previously.67 From the perspective of the
imine mechanism in the acidic environment of material 4, the
protonation state of the amine remains unchanged by addition
of small amounts of water, because the amine is already
protonated due to the acidity of the surface. This protonation
of the amine in material 4 is not a significant barrier to efficient
catalysis due to the overall favorable acid-base cooperativity,
which synthesizes imine rapidly under bifunctional condi-
tions15,56,57 and favors the strongly electrophilic iminium ion
tautomer as observed in our salicylaldehyde binding results. We
further note that the effect of water in the Henry reaction is
self-promoting. Water disfavors the ion-pair mechanism relative
to the imine mechanism, which, unlike the ion-pair mechanism,
produces water during reaction. The presence of small amounts
of water therefore leads to further water synthesis during Henry
catalysis, due to promotion of the imine over ion-pair mecha-
nism.
Henry product ratiob
[TOF]b
a
a
material
Michael kapp
Knoevenagel kapp
4
5
6
0.35
0.006
0.024
2.0
0.026
0.1
85:15 [10]
68:32 [0.6]
60:40 [1.2]
a h-1, using -dCaldehyde/dt ) kappCaldehyde; at low conversion for Michael
reaction. b Ratio of dehydrated to alcohol product (25% aldehyde conver-
sion), TOF: initial number of reaction events h-1 amine site-1
.
immobilized capping agent do not affect catalysis and are
inconsequential for active site accessibility.
The acid-base cooperativity demonstrated in material 4
requires a diverse range of structural requirements between the
bound amine intermediate and the acidic silanols acting on it.
Such versatility is in stark contrast with the specificity of acid-
base distances generally observed in homogeneous sys-
tems.8-10,15,38 Likewise, homogeneous catalysts 7 and 8 did not
show acid-base cooperativity in the form of bifunctional
catalysis, exhibiting catalytic activities comparable to n-butyl-
amine.66 This highlights the important role of the silica surface
in organizing acid-base functional groups for bifunctional
catalysis and shows that the simple presence of acid and base
functions in stoichiometric proportions and close proximity on
the same molecule is insufficient for acid-base cooperativity
in catalysis.
Small amounts of water in the reaction mixture have a
dramatic effect on the observed catalysis proceeding via ion-
pair mechanisms, but a negligible effect on catalysis occurring
via acid-base bifunctional mechanisms. This effect can be
observed simply by using materials that have been stored in
the presence of atmospheric moisture instead of catalysts that
are dehydrated before reaction, and amounts to less than 10
wt % of water for all catalysts as measured by thermogravimetric
analysis, which is less than that required to fill the pores of the
catalyst with water. For materials equilibrated with ambient
moisture before use in Knoevenagel catalysis, the rate enhance-
ment of 4 is 20-fold over 6 and 75-fold over 5. This effect is
not confined to the Knoevenagel reaction, but is general for all
three catalytic reactions. These data are summarized in Table
3. These marked rate and selectivity differences between acidic
and nonacidic catalysts reflect a lower rate on nonacidic
materials upon exposure to water, while the activity of 4 remains
essentially unchanged for all three reactions. This is particularly
Conclusions
We utilize a materials synthesis approach for elucidating the
independent contributions of outer-sphere dielectric environment
and acid-base cooperativity on primary amine reactivity in a
heterogeneous hybrid organic-inorganic catalyst. In the Henry
reaction, these two outer-sphere properties are critical in
determining both the rate and the mechanism of the reaction.
While both polar materials facilitate catalysis by at least 20-
fold over the nonpolar catalyst 5, they show opposite product
selectivity. Because of the rapid formation of the highly
electrophilic iminium cation intermediate, the acid-base bi-
functional material 4 follows a zero-order dependence on
aldehyde reactant concentration on its catalytic rate and is highly
selective for the R,â-unsaturated product 11. In contrast, the
polar/nonacidic material 6 facilitates catalysis through a non-
covalent ion-pair mechanism due to its high dielectric environ-
ment and the lack of acidity in the outer sphere. This translates
into a measured selectivity of 99% for the â-nitro alcohol
product 12. The synthesis of an active primary amine catalyst
that facilitates products via an ion-pair mechanism selectively
over imine-derived products has not been observed for primary
amine catalysts previously. Supporting evidence for the mech-
anisms above is provided by solid-state UV/vis spectroscopy
on salicylaldehyde treated catalysts 4-6. The spectrum of
salicylaldehyde-treated bifunctional catalyst 4 contains a pre-
dominant band that is due to the strong thermodynamic
preference for forming iminium zwitterionic tautomer in this
(66) For the Knoevenagel reaction, kapp for 7 and n-butylamine was less than
(67) Allen, G. C.; Sorbello, F.; Altavilla, C.; Castorina, A.; Ciliberto, E. Thin
Solid Films 2005, 483, 306-311.
0.01 h-1, using -dCaldehyde/dt ) kappCaldehyde
.
9
J. AM. CHEM. SOC. VOL. 128, NO. 11, 2006 3745

A R T I C L E S
Bass et al.
1
Conversion and selectivity were determined via H NMR by taking
aliquots via syringe filter and diluting with CDCl3. Resonances: 5.6
ppm (â-nitro alcohol product), 7.5 ppm (nitrobenzene), 7.72-7.74 ppm
(R,â-unsaturated product), and 10.1 ppm (4-nitrobenzaldehyde).
Knoevenagel Condensation. A typical reaction was conducted using
about 10 mg of catalyst (the amount of catalyst was fixed at 0.01 molar
equivalents of amine relative to 3-nitrobenzaldyde) in 8 mL of an
anhydrous benzene solution of concentration 0.022 M in 3-nitroben-
zaldyde and 0.044 M of malononitrile. To capped catalysts was added
mesoporous silica (Selecto) as a minimally catalytic desiccant in a ratio
that kept the total number of acidic silanols in the reaction constant.
Catalysts were prepared via thermolysis (Supporting Information) in
the reactor vessel (volume was 10 mL) prior to reactant addition and
kept air free under a N2 environment. The reaction was performed at
room temperature (22 °C), and aliquots were taken by syringe and
analyzed by gas chromatography using 1,3,5-trimethoxybenzene as an
internal standard.
material, and a minor band due to neutral imine tautomer. In
general catalysts 5 and 6 after salicylaldehyde treatment, there
is reversal in the order of intensity of these two bands.
Outer-sphere environment effects are further manifested in
carbamate thermolysis, where the acid-base bifunctional mate-
rial lowers the activation barrier by 2.7 kcal/mol while only a
small dielectric effect of 0.5 kcal/mol is observed. Catalysis in
the Michael and Knoevenagel reactions echo the Henry results
with the high dielectric materials being more active than the
nonpolar material 5. In the presence of water, however, only
the acid-base bifunctional material 4 performs well compared
to anhydrous conditions, resulting in an order of magnitude or
higher activity in comparison with 5 and 6 for all catalytic
reactions. This is due to the depressed catalytic rate in the
nonacidic materials for the ion-pair mechanism, as manifested
in changes in the Henry reaction selectivity, likely due to
competitive protonation by water. The remarkably diverse set
of reactions affected by acid-base bifunctional cooperativity
observed in this Article, including the various catalytic probe
reactions, zwitterion formation during salicyladehyde binding,
and immobilized carbamate thermolysis, is testament to the
promiscuity of acid-base distances attainable in heterogeneous
hybrid organic-inorganic catalysts. These data prove the
concept hypothesized in Scheme 1, relating to the silica surface
providing a range of acid-base distances for bifunctional
cooperativity. In contrast, no evidence was observed for
bifunctional cooperativity in homogeneous calixarene models
despite the presence of acidic groups juxtaposed to a pendent
amine. Altogether, our results are significant for understanding
acid-base bifunctional reactivity in heterogeneous organic-
inorganic catalysts, as well as the chemical reactivity of
heterogeneous primary amine catalysts.
Synthesis of NBoc Protected Calixarenes 7 and 8 via Mitsunobu
Reaction. To a mixture of calix[4]arene (0.154 mmol) and NBoc-
alcohol (0.23 mmol) in 5 mL of THF was added triphenyl phosphine
(0.23 mmol). The resulting mixture was cooled to 0 °C, and diethyl-
azodicarboxylate (0.23 mmol) as 40% solution in toluene was added
dropwise. After 10 min of stirring, the resulting yellow-green solution
was allowed to warm to room temperature. After 24 h of stirring at
room temperature, the solution was evaporated to dryness and the
residue was treated with 2 mL of methanol for 20 min followed by
evaporation. Purification via column chromatography yielded the final
products.
5,11,17,23-Tetra-(tert-butyl)-25-[3-(tert-Butoxycarbonylamino)-
propoxy]-calix[4]arene-26,27,28-triol (Nboc7). After separation by
column chromatography (CH2Cl2/methanol ) 1:0.025 v/v, Rf 0.4), white
powder was obtained in 55% yield: 1H NMR (400 MHz) δ 10.24 (s,
1H, OH), 9.67 (s, 2H, OH), 7.13 (s, 2H, ArH), 7.12 (d, 2H, 2.0 Hz,
ArH), 7.10 (s, 2H, ArH), 7.04 (d, 2H, 2.4 Hz, ArH), 5.80 (s, 1H, NH),
4.31, 4.33 (two d, 2H+2H, 13.2 Hz, ArCH2Ar), 4.19 (t, 2H, 6.0 Hz,
OCH2), 3.64 (m, 2H, CH2N), 3.49 (d, 4H, 13.2 Hz, ArCH2Ar), 2.35
(m, 2H, OCH2CH2), 1.52 (s, 9H, OC(CH3)3), 1.27 (s, 9H, C(CH3)3),
1.26 (s, 18H, C(CH3)3), 1.23 (s, 9H, C(CH3)3).
25-[4-(tert-Butoxycarbonylamino)butoxy]-calix[4]arene-26,27,28-
triol (Nboc8). After separation by column chromatography (CH2Cl2/
methanol ) 1:0.02 v/v, Rf 0.6), white powder was obtained in 50%
yield: 1H NMR (400 MHz) δ 9.74 (s, 1H, OH), 9.72 (s, 2H, OH),
7.04, 7.06, 7.10, 7.12 (four d, 2H+2H+2H+2H, 7.6 Hz, ArH-m), 6.91
(t, 1H, 7.6 Hz, ArH-p), 6.72 (t, 1H, 7.6 Hz, ArH-p), 6.71 (t, 2H, 7.6
Hz, ArH-p), 4.85 (s, 1H, NH), 4.31, 4.38 (two d, 2H+2H, 12.8 Hz,
ArCH2Ar), 4.20 (t, 2H, 6.8 Hz, OCH2), 3.50, 3.51 (two d, 2H+2H,
12.8 Hz, ArCH2Ar), 3.40 (m, 2H, CH2N), 2.21 (m, 2H, OCH2CH2),
1.97 (m, 2H, CH2CH2N), 1.57 (s, 9H, OC(CH3)3).
Experimental Section
General. All reagents were purchased from Aldrich at analytical
grade and used as received, unless otherwise noted. 3-Cyanopropyl-
dimethylchlorosilane was distilled prior to use. Dry solvents were
prepared via distillation using standard methods. 1H NMR spectra were
recorded in CDCl3 (293 K) either on a Bruker AV-300 (300 MHz)
instrument or on an AVB-400 (400 MHz) instrument. Single-crystal
diffraction data were collected at the UC Berkeley College of Chemistry
X-ray Crystallographic Facility. All catalytic reactions were conducted
in a well-ventilated fume hood. FAB-MS spectra were recorded using
a o-nitrophenyl octyl ether (NPOE) matrix at the UC Berkeley Mass
Spectrometry Facility.
Deprotection of Nboc-Calixarenes. To a solution of NBoc-protected
calixarenes (0.09 mmol) in 3.0 mL of dry dichloromethane was added
trifluoroacetic acid (0.9 mmol) at 0 °C. The solution was allowed to
warm to room temperature and then stirred an additional 10 h. The
reaction mixture was then hydrolyzed using a saturated NaHCO3
solution. The aqueous solution was extracted with CH2Cl2 and dried
over Na2SO4. After evaporation of solvent, residue was washed with
hexane to give desired calixarene-amines 7 and 8.
5,11,17,23-Tetra-(tert-butyl)-25-[3-aminopropoxy]-calix[4]arene-
26,27,28-triol (7). White powder was obtained in 98% yield: 1H NMR
(400 MHz) δ 9.00 (br s, 3H+2H, OH+NH2), 7.07 (s, 2H, ArH), 7.00
(s, 4H, ArH), 6.97 (s, 2H, ArH), 4.13 (m, 4H+2H, ArCH2Ar+OCH2),
3.51 (m, 2H, OCH2N), 3.35, 3.39 (two d, 13.6 Hz, ArCH2Ar), 2.43
(m, 2H, OCH2CH2), 1.23 (s, 18H, C(CH3)3), 1.20 (s, 9H, C(CH3)3),
1.13 (s, 9H, C(CH3)3). FAB MS m/z 706.6 [M + H+].
Michael Addition. A typical reaction was conducted using about
10 mg of catalyst (the amount of catalyst was fixed at 0.02 molar
equivalents of amine relative to trans-â-nitrostyrene) in 8 mL of an
anhydrous benzene solution of concentration 0.022 M in trans-â-
nitrostyrene and 0.044 M of malononitrile. Catalysts were prepared
via thermolysis (Supporting Information) in the reactor vessel (volume
was 10 mL) prior to addition and kept air free under a N2 environment.
The reaction was performed at room temperature (22 °C), and aliquots
were taken by syringe and analyzed by gas chromatography using 1,3,5-
trimethoxybenzene as an internal standard.
Henry Reaction. Reactions were conducted using approximately
30 mg of catalyst (the amount of catalyst was fixed at 0.01 molar
equivalents of amine relative to 4-nitrobenzaldehyde). Catalysts were
prepared via thermolysis (Supporting Information) in the reactor vessel
(volume was 10 mL) prior to addition and kept air free under a N2
environment. To this vessel was added a solution containing 5 mmol
nitromethane, 0.5 mmol 4-nitrobenzaldehyde, and 0.05 mmol nitroben-
zene as an internal standard. The reaction was performed at 40 °C.
25-[4-Aminobutoxy]-calix[4]arene-26,27,28-triol (8). White pow-
der was obtained in 95% yield: 1H NMR (400 MHz) δ 7.50 (br s,
3H+2H, OH+NH2), 6.94-7.06 (m, 8H, ArH), 6.58-6.83 (m, 4H,
9
3746 J. AM. CHEM. SOC. VOL. 128, NO. 11, 2006

Outer-Sphere Effects in Heterogeneous Catalysis
A R T I C L E S
ArH), 4.24, 4.28 (two d, 2H+2H, 13.2 Hz, ArCH2Ar), 3.99 (t, 2H, 7.6
Hz, OCH2), 3.42 (d, 4H, 13.2 Hz, ArCH2Ar), 2.99 (t, 2H, 7.6 Hz,
CH2N), 2.08 (m, 2H, OCH2CH2), 1.87 (m, 2H, NCH2CH2). MS ES
m/z 496.3 [M + H+].
m, 8H, ArH), 6.92 (m, 2H, C6H4), 6.72 (m, 4H, ArH), 4.29, 4.38 (two
d, 2H+2H, 13.5 Hz, ArCH2Ar), 4.24 (m, 2H, OCH2), 3.90 (m, 2H,
NCH2), 3.51 (d, 4H, 13.5 Hz, ArCH2Ar), 2.27 (m, 4H, OCH2-
CH2+NCH2CH2).
General Procedure for Preparation of Calixarene-Imines (9,10).
The solution of calixarene amine 7 (or 8) (0.014 mmol) and salicyla-
ldehyde (0.017 mmol) in 2 mL of m-xylene was refluxed over molecular
sieves for 10 h. The organic solvent was evaporated, and the residue
was washed with hexane. Hexane was separated, and the precipitate
formed was dried in a vacuum (0.05 Torr) for 3 h.
5,11,17,23-Tetra-(tert-butyl)-25-[3-(2-hydroxyphenylmethylen-
imino)propoxy]-calix[4]arene-26,27,28-triol (9). Yellow powder was
obtained in 98% yield: 1H NMR (400 MHz) δ 14.20 (s, 1H,
o-HOC6H4), 10.24 (s, 1H, OH), 9.64 (s, 2H, OH), 8.73 (s, 1H, CHN),
7.18, 7.35, 7.60 (three m, 1H each, C6H4), 7.05-7.15 (br m, 8H, ArH),
6.92 (m, 1H, C6H4), 4.33, 4.36 (two d, 2H + 2H, 13.6 Hz, ArCH2Ar),
4.23 (m, 4H, OCH2+NCH2), 3.47, 3.51 (two d, 2H+2H, 13.6 Hz,
ArCH2Ar), 2.57 (m, 2H, OCH2CH2), 1.28 (s, 9H, C(CH3)3), 1.26 (s,
18H, C(CH3)3), 1.23 (s, 9H, C(CH3)3). FAB MS m/z 810.5 [M+H+].
25-[4-(2-Hydroxyphenylmethylenimino)-propoxy]-calix[4]arene-
26,27,28-triol (10). Yellow powder was obtained in 96% yield: 1H
NMR (400 MHz) δ 13.44 (s, 1H, o-HOC6H4), 9.72 (s, 1H, OH), 9.40
(s, 2H, OH), 8.53 (s, 1H, CHN), 7.33 (m, 2H, C6H4), 7.00-7.15 (br
Acknowledgment. Funding from the National Science Foun-
dation is gratefully acknowledged (CTS 0407478). A.J.P. is
grateful to the Robert and Colleen Haas Scholars Program for
a fellowship. The authors thank Dr. Frederick Hollander and
Dr. Allen Oliver of the UC Berkeley CHEXRAY facility for
determining the structure of 7 via single-crystal X-ray diffrac-
tion; Mr. Justin M. Notestein for technical assistance with FTIR
spectroscopy; and Ms. Karen Magid for technical assistance with
SEM.
Supporting Information Available: Supplementary figures
referenced in the text, silica synthesis, capping procedures,
thermolysis conditions, salicylaldehyde binding procedures,
activation energy measurements, and X-ray crystallographic
information (CIF). This material is available free of charge via
the Internet at http://pubs.acs.org.
JA057395C
9
J. AM. CHEM. SOC. VOL. 128, NO. 11, 2006 3747