Kinetic and Mechanistic Examination of Acid-Base Bifunctional Aminosilica Catalysts in Aldol and Nitroaldol Condensations

Article abstract of DOI:10.1021/acscatal.5b02398

The kinetic and mechanistic understanding of cooperatively catalyzed aldol and nitroaldol condensations is probed using a series of mesoporous silicas functionalized with aminosilanes to provide bifunctional acid-base character. Mechanistically, a Hammett analysis is performed to determine the effects of electron-donating and electron-withdrawing groups of para-substituted benzaldehyde derivatives on the catalytic activity of each condensation reaction. This information is also used to discuss the validity of previously proposed catalytic mechanisms and to propose a revised mechanism with plausible reaction intermediates. For both reactions, electron-withdrawing groups increase the observed rates of reaction, though resonance effects play an important, yet subtle, role in the nitroaldol condensation, in which a p-methoxy electron-donating group is also able to stabilize the proposed carbocation intermediate. Additionally, activation energies and pre-exponential factors are calculated via the Arrhenius analysis of two catalysts with similar amine loadings: one catalyst had silanols available for cooperative interactions (acid-base catalysis), while the other was treated with a silanol-capping reagent to prevent such cooperativity (base-only catalysis). The values obtained for activation energies and pre-exponential factors in each reaction are discussed in the context of the proposed mechanisms and the importance of cooperative interactions in each reaction. The catalytic activity decreases for all reactions when the silanols are capped with trimethylsilyl groups, and higher temperatures are required to make accurate rate measurements, emphasizing the vital role the weakly acidic silanols play in the catalytic cycles. The results indicate that loss of acid sites is more detrimental to the catalytic activity of the aldol condensation than the nitroaldol condensation, as evidenced by the significant decrease in the pre-exponential factor for the aldol condensation when silanols are unavailable for cooperative interactions. Cooperative catalysis is evidenced by significant changes in the pre-exponential factor, rather than the activation energy for the aldol condensation.

Full text of DOI:10.1021/acscatal.5b02398

This is an open access article published under an ACS AuthorChoice License, which permits
copying and redistribution of the article or any adaptations for non-commercial purposes.
Research Article
pubs.acs.org/acscatalysis
Kinetic and Mechanistic Examination of Acid−Base Bifunctional
Aminosilica Catalysts in Aldol and Nitroaldol Condensations
Virginia E. Collier, Nathan C. Ellebracht, George I. Lindy, Eric G. Moschetta,* and Christopher W. Jones*
School of Chemical & Biomolecular Engineering Georgia Institute of Technology 311 Ferst Drive NW, Atlanta, Georgia 30332,
United States
*
S Supporting Information
ABSTRACT: The kinetic and mechanistic understanding of
cooperatively catalyzed aldol and nitroaldol condensations is probed
using a series of mesoporous silicas functionalized with aminosilanes
to provide bifunctional acid−base character. Mechanistically, a
Hammett analysis is performed to determine the effects of electron-
donating and electron-withdrawing groups of para-substituted
benzaldehyde derivatives on the catalytic activity of each con-
densation reaction. This information is also used to discuss the
validity of previously proposed catalytic mechanisms and to propose a
revised mechanism with plausible reaction intermediates. For both reactions, electron-withdrawing groups increase the observed
rates of reaction, though resonance effects play an important, yet subtle, role in the nitroaldol condensation, in which a p-
methoxy electron-donating group is also able to stabilize the proposed carbocation intermediate. Additionally, activation energies
and pre-exponential factors are calculated via the Arrhenius analysis of two catalysts with similar amine loadings: one catalyst had
silanols available for cooperative interactions (acid−base catalysis), while the other was treated with a silanol-capping reagent to
prevent such cooperativity (base-only catalysis). The values obtained for activation energies and pre-exponential factors in each
reaction are discussed in the context of the proposed mechanisms and the importance of cooperative interactions in each
reaction. The catalytic activity decreases for all reactions when the silanols are capped with trimethylsilyl groups, and higher
temperatures are required to make accurate rate measurements, emphasizing the vital role the weakly acidic silanols play in the
catalytic cycles. The results indicate that loss of acid sites is more detrimental to the catalytic activity of the aldol condensation
than the nitroaldol condensation, as evidenced by the significant decrease in the pre-exponential factor for the aldol condensation
when silanols are unavailable for cooperative interactions. Cooperative catalysis is evidenced by significant changes in the pre-
exponential factor, rather than the activation energy for the aldol condensation.
KEYWORDS: cooperative catalysis, bifunctional acid−base catalysts, Hammett analysis, Arrhenius analysis, condensation reactions,
carbon−carbon coupling
INTRODUCTION
the surfaces of bifunctional catalysts often must be kept spatially
separated such that they do not negate each other’s effects (i.e.,
acids and bases), as they would in solution, while still allowing
them to interact favorably. Mesoporous silica is a popular
choice for a support because the surface silanols act as weak
■
In organic synthesis, the formation of carbon−carbon bonds is
an essential step in the production of complex molecules in the
fine chemical and pharmaceutical industries. These reactions
often require homogeneous transition-metal catalysts, which are
expensive, can deactivate and/or deteriorate over time, and are
difficult to separate and recycle from the reaction mixture.
Silica-tethered organocatalysts are chemically customizable, are
easily separated, and do not require transition metals for several
7
1
−3
Brønsted acids, making them viable as acid partners in acid−
10−27
4
−6
base cooperative catalysts.
Additionally, the range of
alkoxysilanes that can be grafted to the native silanols allows for
a highly tunable series of organic moieties that can be tethered
10,11,22,27
7
to the surface to catalyze the desired reaction.
A
important condensation reactions. The design of these
further understanding of these materials has led to practical
developments using supported organocatalysts, such as
upgrading biomass-derived molecules to materials that may
be used as alternative fuel sources, which illustrates their
promise as catalysts for important chemical transforma-
organocatalytic mesoporous silicates is inspired by enzymatic
catalysts, in which the functional groups contained in the
overall enzyme structure are carefully spaced apart and conform
to the substrate, stabilizing the transition state and reducing the
8
,9
required activation energy.
Heterogeneous bifunctional
17,28
tions.
catalysts, materials that contain two disparate catalytic entities,
often act as cooperative catalysts, by which the two species
operate in concert to increase the overall rate of reaction in
such a fashion that would not be achievable using either of the
Received: October 25, 2015
Revised: December 8, 2015
7
catalytic species independently. The active species found on
©
XXXX American Chemical Society
460
DOI: 10.1021/acscatal.5b02398
ACS Catal. 2016, 6, 460−468

ACS Catalysis
Research Article
Previous work on cooperative acid−base catalysts inves-
tigated the physical (e.g., alkyl linker length, pore diameter,
arrangement of acids and bases) and chemical properties (e.g.,
types and strengths of acid and base sites) of such materials to
understand the fundamental interactions occurring in the
catalytic mechanisms. Brunelli et al. found that aminopropyl
the aldehyde with subsequent nucleophilic attack of the nitro-
containing molecule.
32−35
In spite of these advances in the design of heterogeneous
cooperative aminosilica catalysts, there are still important areas
that must be explored to provide key information needed for
future catalyst and reaction design. While the above studies
have elucidated some key intermediates and offered suggestions
for improving bifunctional catalyst design, little has been
reported regarding the activation energies and pre-exponential
factors of the aldol and nitroaldol condensations using acid−
(C3) linkers offered high activity (turnover frequency, TOF)
for the aldol condensation compared to alkyl linkers with fewer
methylene units, while increasing the alkyl linker length past C3
23
did not increase the observed activity. Later work showed that
as the pore diameter of the silica changed, the required alkyl
linker length to achieve the maximum TOF for aldol and
24
base catalysts. Such information would provide critical
information regarding (i) a more quantifiable way to assess
the importance of the availability of the silanols as acid partners
for the amines via changes in the pre-exponential factors and
24
nitroaldol reactions also changed. Lauwaert et al. examined
the effect of the silanol to amine ratio for the aldol
condensation and obtained the maximum TOF for a ratio of
(
ii) possible effects on the transition state by examining changes
2
5
in the activation energies in each reaction. Furthermore,
calculating the activation energies and pre-exponential factors
will reveal if molecular collisions (pre-exponential factor) or
surpassing a reaction activation barrier has the more significant
effect in these catalytic reactions. To this end, in this work, the
physical chemistry of the aldol and nitroaldol condensations is
explored using MCM-41, a mesoporous silica, functionalized
with (4-aminobutyl)triethoxysilane. First, Hammett analysis
was performed using the same catalyst for both the aldol and
nitroaldol condensations to determine the effects of electron-
donating and electron-withdrawing groups on the rates of
reaction and the nature of potential reaction intermediates.
Additionally, Arrhenius analysis was performed for the aldol
and nitroaldol condensations to calculate the observed
activation energies and pre-exponential factors. This analysis
was repeated for a catalyst at a similar amine loading that
contained trimethylsilyl-capped silanols, reducing the cooper-
ative interactions of the silanols with the amines. Comparing
the activation energies and pre-exponential factors for both
condensation reactions using the catalysts with uncapped and
capped silanols has provided quantitative insight into the
mechanisms of the cooperative catalytic reactions. Such
information can be useful in tailoring the design of future
cooperative catalysts.
1
.7. Sharma et al. studied the effect of the total aminosilane
loading on MCM-41 silica for the nitroaldol condensation and
found that TOFs were highest for loadings between 0.8 and 1.5
29
mmol of N/g of silica. Another study demonstrated that
spacing between silanols and secondary amines on SBA-15
materials could be tuned by changing the calcination
temperature of the silica, which affected the concentration
16
and acid strength of the silanols. Several studies have
examined the impact of acid strength and spatial organization
on cooperative catalysis, noting that weaker acids such as
silanols were best at promoting acid−base cooperativity with
7
,17,22,27,30
primary amines.
Moschetta et al. examined the effect
of incorporating heteroatoms into the silica framework and
showed that slight increases in the number of Lewis acid sites
increased the catalytic activity for the nitroaldol condensation
30
but not the aldol condensation. Shylesh et al. studied the gas-
phase self-condensation of butanal over heteroatom-doped
SBA-15 silicas functionalized with secondary amines and
observed increased catalytic activity for samples containing
14
stronger Brønsted acid sites. These studies suggest that the
optimal acid strength depends on the nature of the amine
grafted on the surface, as well as the specifics of the coupling
reaction, with aldol and nitroaldol reactions showing different
trends.
However, additional challenges remain in understanding the
mechanisms operating in these condensation reactions using
bifunctional catalysts. For example, the aldol and nitroaldol
condensations are often used to study the activity of acid−base
catalysts, but such studies typically consider one reaction or the
EXPERIMENTAL SECTION
Materials. Acetone (99.5%), hexane (98.5%), and toluene
99.5%) were purchased from BDH. The aminosilane (4-
aminobutyl)triethoxysilane (95%) was purchased from Gelest,
-nitrobenzaldehyde (99%) and ammonia (28 wt %) were
purchased from Alfa Aesar, 1,4-dimethoxybenzene (99%) was
procured from Tokyo Chemical Industry (TCI), and
cetyltrimethylammonium bromide (CTAB) (99%) was pur-
chased from Acros. Nitromethane (95%), tetraethyl orthosili-
cate (TEOS; 98%), 4-bromobenzaldehyde (99%), benzalde-
hyde (98%), 4-methoxybenzaldehyde (98%), 4-methylbenzal-
dehyde (97%), and hexamethyldisilazane (HMDS, 99.9%) were
purchased from Sigma-Aldrich. Ethanol (100%) was obtained
from Koptec, and CDCl3 (99.8%) was obtained from
Cambridge Isotope Laboratories, Inc.
■
(
4
20,25,29
other, not both, which makes cross-comparisons difficult.
Kandel et al. spectroscopically verified the formation of stable
imines on primary amine-functionalized silica surfaces in the
presence of aldol condensation substrates, confirming the role
of imines, specifically aldimines resulting from amines reacting
with aldehydes, as off-cycle species that inhibited the observed
3
1
rates of reaction. Another study examined different types of
secondary amines on propylsilanes, noting that only a
methylamine catalyst had an increased activity in the aldol
26
condensation in comparison to the primary amine. For the
aldol condensation, Brunelli et al. postulated sequential
activation of the ketone (electrophile) to form the reactive
enamine intermediate using a single silanol, which then attacks
Synthesis of MCM-41. MCM-41 was synthesized accord-
24
ing to a previous report. Cetyltrimethylammonium bromide
(7.6 mmol) was dissolved in distilled water (121 mL) with an
aqueous ammonia solution (28 wt %, 8.5 g) in a 250 mL flask at
room temperature. The mixture was stirred vigorously. TEOS
(10 g) was added dropwise, and the mixture was stirred for 2 h.
The mixture was filtered and washed several times with distilled
water, and the powder was dried at 75 °C in an oven overnight.
2
3
the aldehyde. One proposed mechanism of the nitroaldol
condensation involves simultaneous activation of the reactants
24
and uses multiple silanols. Other studies have proposed
mechanisms involving imine intermediates resulting from
nucleophilic attack by the amine on the carbonyl group of
4
61
DOI: 10.1021/acscatal.5b02398
ACS Catal. 2016, 6, 460−468

ACS Catalysis
Research Article
The powder was calcined under a flow of air, ramped from
room temperature to 550 °C at 1.2 °C/min, held at 550 °C for
reactant and 0.05 M 1,4-dimethoxybenzene as an internal
standard in acetone was added to a clean 25 mL two-neck flask
with a magnetic stir bar. Next, 10 mol % of catalyst (by amine
loading) was added to the reaction solution and the reaction
was run in an oil bath at 50 °C under a condenser with an
atmosphere of nitrogen. Samples were taken every 15 min for
the first 1 h and then one sample per hour for a total of 4 h. All
reactions were performed three times to ensure reproducibility,
and the average conversion values of the reactions were used to
calculate TOFs during the first 1 h of reaction (the conversion
data as a function of time were linear in this region). For each
nitroaldol condensation, 1 mL of reaction solution containing
0.05 M 4-nitrobenzaldehyde as reactant and 0.05 M 1,4-
dimethoxybenzene as an internal standard in nitromethane was
placed in a clean 25 mL two-neck flask with a magnetic stir bar.
Next, 10 mol % of catalyst (by amine loading) was added to the
reaction solution and the reaction was run in an oil bath at 40
°C under a condenser with an atmosphere of nitrogen. Samples
were taken every 15 min for the first 1 h and then one sample
per hour for a total of 4 h. All reactions were performed three
times to ensure reproducibility, and the average conversion
values of the reactions were used to calculate TOFs during the
first 1 h of reaction (once again, the conversion data as a
function of time were linear). Reactant conversions were
determined by GC (Shimadzu GC-2010) with a flame
ionization detector (FID), an autoinjector and autosampler
12 h, and then cooled to room temperature (10 °C/min). The
powder was dried overnight at 200 °C under reduced pressure
(
10 mTorr).
Grafting of (4-Aminobutyl)triethoxysilane on MCM-
4
1. Silica samples were dried in 100 mL flasks under vacuum at
100 °C overnight. Under nitrogen pressure, the desired amount
of (4-aminobutyl)triethoxysilane was added to 500 mg of dried
silica in 12.5 mL of anhydrous toluene. Silica samples were
stirred at room temperature for 24 h and then heated to 80 °C
in an oil bath and reacted for 24 h. The samples were washed,
filtered with 100 mL each of toluene, hexane, and ethanol, and
then dried under vacuum at 100 °C overnight. The amine
loading was measured by elemental analysis (EA), and the
results are given in Table 1. In this paper, the naming scheme
Table 1. Textural Properties of the Parent MCM-41 Silica
and the Resulting Aminosilica Bifunctional Catalysts
surface
area
total pore
volume
pore
amine loading
diameter
(mmol of N/
2
a
3
a
a
b
material
MCM-41
MCM-C4-
(m /g)
(cm /g)
(nm)
g of silica)
1109
970
0.97
0.78
2.9
2.7
0.44
0.57
0
.44-H
MCM-C4-
.57-A
MCM-C4-
.57-A-HMDS
965
802
0.76
0.60
2.7
2.4
0
(
AOC-20i+s model), and an Agilent DB-1701 column (30 m,
0.57
0
.25 mm inner diameter, 0.25 μm film thickness).
0
a
b
Hammett Analysis. The Hammett reaction setup was
Determined by nitrogen physisorption experiments at 77 K. Amine
replicated as detailed above using MCM-C4-0.44-H and one of
five different para-substituted benzaldehyde derivatives (4-
nitrobenzaldehyde, 4-methoxybenzaldehyde, 4-bromobenzalde-
hyde, 4-methylbenzaldehyde, and benzaldehyde). Samples for
each new reactant were taken every 15 min for 1 h to calculate
the TOFs. Each derivative reaction was run three times, and
average TOFs are reported. All Hammett substituent constants
loadings determined by elemental analysis.
for the aminosilica catalysts follows the convention “MCM-C4-
n-X”, where C4 refers to the length of the aminobutyl linker, n
is the amine loading in mmol of N/g of silica, and X refers to
whether the catalyst was used in the Hammett (H) or
Arrhenius analysis (A). The catalyst that was treated with
HMDS has “HMDS” at the end of its name.
38
were taken from a previous review of reported values. The
crude reaction mixtures of the nitroaldol condensation were
1
Materials Characterization. All nitrogen physisorption
experiments were conducted on a Micromeritics Tristar 2030
instrument at 77 K. All samples were evacuated for at least 12 h
at 110 °C using at least 100 mg of each material. The nitrogen
uptake isotherms were analyzed using the Brunauer−Emmett−
Teller (BET) method to calculate the surface area of each
material and the Broekhoff−de Boer method with the Frenkel−
Halsey−Hill (BdB-FHH) modification using the data from the
adsorption isotherm to calculate the cumulative pore volume
examined by H NMR spectroscopy to determine the
selectivity, though only the nitrostyrene products were present,
34
in agreement with a previous report.
Arrhenius Analysis. The aldol and nitroaldol condensa-
tions were performed as detailed above using MCM-C4-0.57-A
as the catalyst and 4-nitrobenzaldehyde as the reactant, varying
only the reaction temperature. The aldol reaction was run from
45 to 55 °C, while the nitroaldol condensation was run from 40
to 55 °C. Each set of reactions was performed again using an
HMDS-treated catalyst (MCM-C4-0.57-A-HMDS) with a
similar amine loading. A portion of MCM-C4-0.57-A catalyst
was capped by vigorous mixing with neat HMDS and allowed
to react for 24 h at room temperature, resulting in a catalyst
with capped silanols. The HMDS-treated catalyst was washed
with 100 mL each of toluene, hexanes, and ethanol and dried
under vacuum at 100 °C overnight. Due to the presence of the
capped silanols, the reaction temperatures were increased for
the HMDS-treated catalyst. Four reactions were run in parallel
in 10 mL pressure tubes. Each vessel was loaded with a
magnetic stir bar, 2.7 mL of reaction solution, and 5 mol % (by
amine loading) of capped MCM-C4-0.57-A-HMDS catalyst.
The reactions were carried out in pressure tubes to prevent
evaporation of solvent from the reaction solution over time, as
they required elevated temperatures and increased reaction
times to observe measurable conversion values. An aliquot was
36
and the pore diameter. X-ray diffraction (XRD) experiments
were performed on a PAnalytical X’Pert diffractometer using a
Cu Kα X-ray source. Samples were scanned at a grazing angle
of 1° and a step size of 0.02° over a 2θ range of 1−10°. All
samples were sent to Atlantic Microlabs (Norcross, GA) for
1
CHN elemental analysis (EA). All H NMR spectra of the
crude mixtures of the nitroaldol condensation products were
recorded on a Bruker DMX 400 spectrometer using CDCl as
3
the solvent. No dinitroalkane products were observed in any of
3
7
the crude mixtures. Transmission electron microscopy
TEM) of the MCM-41 can be found in Figure S6 of the
Supporting Information, and other microscopy images can be
(
24
found in a previous report.
General Catalytic Experiments for MCM-41 Materials
with Uncapped Silanols. For each aldol condensation, 2 mL
of reaction solution containing 0.05 M 4-nitrobenzaldehyde as
4
62
DOI: 10.1021/acscatal.5b02398
ACS Catal. 2016, 6, 460−468

ACS Catalysis
Research Article
taken from each pressure tube prior to heating as the initial
time point. For each subsequent time point (every 30 min for
nitroaldol, 1 h for aldol), the pressure tubes were removed from
the oil bath and cooled in an ice bath for 5 min to reduce
solvent vapor escaping the tube upon opening. During this
time, the reaction timer was stopped. After aliquots were taken
from each vessel, they were placed back in the oil bath and the
reaction timer was started after a short delay to allow for
temperature equilibration. All conversion data used to calculate
TOFs, and therefore activation energies and pre-exponential
factors, were linear as a function of time.
RESULTS AND DISCUSSION
■
Physical and Chemical Characterization of Materials.
Nitrogen physisorption was used to characterize the textural
properties of the MCM-41 materials, the results of which are
reported in Table 1. The adsorption isotherms (Figure 1)
Figure 2. XRD patterns of the bare MCM-41 silica and the amine-
functionalized MCM-41 silicas, showing the (100), (110), and (200)
reflections in all cases.
with increased aminosilane loadings had decreased surface
areas, pore volumes, and pore diameters, which is consistent
with the filling the pores of the MCM-41 silica with the
aminosilanes.
Hammett Analysis of Cooperative Catalysts. The
reaction schemes for the aldol and nitroaldol condensations
are presented in Scheme 1, and the results of the Hammett
analysis for the aldol condensation are presented in Figure 3.
The TOF values for the five benzaldehyde derivatives were best
fit using the standard σ values, the substituent constants, with a
slope of 0.78, indicating a moderate substituent effect on the
38
activity of the reaction. The positive value of the slope
revealed that electron-withdrawing groups at the para position
on the benzaldehyde accelerated the rate of reaction in the case
Figure 1. Nitrogen physisorption isotherms of the bare MCM-41 silica
and the amine-functionalized MCM-41 silicas used as cooperative
catalysts.
38
of the aldol condensation. Including electron-withdrawing
groups at the para position decreases the electron density at the
carbonyl group of the aldehyde, increasing its electrophilicity
and making the aldehyde more susceptible to nucleophilic
attack. This observation is consistent with the proposed
mechanism of the aldol condensation, illustrated in Scheme
revealed the retention of mesoporosity after functionalization
with the aminosilanes, indicating the grafting conditions were
sufficiently mild. All aminosilica materials had decreased surface
areas, pore volumes, and pore diameters in comparison to the
parent MCM-41 due to the presence of the grafted amino-
silanes. All catalysts were derived from this parent MCM-41 to
ensure the materials had similar textural properties, allowing for
straightforward comparison of their catalytic behavior. The
aminosilica that was treated with HMDS to cap the surface
silanols with trimethylsilyl groups (MCM-C4-0.57-A-HMDS)
exhibited a decrease in surface area, pore volume, and pore
diameter in comparison to the same aminosilica material
23,25,26
2.
The key step in the mechanism is the nucleophilic
attack of the enamine, generated by activation of the acetone,
2
3,25,26
on the aldehyde interacting with a surface silanol.
This
step provides a reasonable explanation for why increasing the
electrophilicity of the aldehyde via incorporation of electron-
withdrawing groups led to increased TOF values in the aldol
condensation.
Conversely, the Hammett analysis for the nitroaldol
condensation revealed interesting differences in comparison
to the aldol condensation, as shown in Figure 4. First, the slope
of the fit is also positive, but the magnitude of the slope is
reduced in comparison to the aldol condensation. This
decreased slope would seem to suggest that substituent effects
do not have a significant effect on the observed kinetics of the
nitroaldol condensation, but closer inspection reveals a subtle
stabilization effect based on the electronic nature of the
substituent. Second, the results of the nitroaldol condensation+
using the five benzaldehyde derivatives were best fit using σ
values, which have different physical meanings in comparison to
traditional σ values (see Figure S1 in the Supporting
(
MCM-C4-0.57-A) without HMDS treatment, an expected
consequence of successful capping of the silanols.
X-ray diffraction (XRD) patterns were obtained for all
MCM-41 materials and the patterns are shown in Figure 2. The
(
100) reflections appeared at 2θ = 2.3−2.5° for all samples, and
the (110) and (200) reflections appeared at 2θ = 4.1−4.8°,
indicating that the 2D hexagonal structure of the parent MCM-
4
1 silica was initially achieved and remained intact on the other
materials after grafting and HMDS treatment, where appro-
priate.
3
9,40
The nitrogen content of each catalyst was
determined by EA and is summarized in Table 1. Samples
4
63
DOI: 10.1021/acscatal.5b02398
ACS Catal. 2016, 6, 460−468

ACS Catalysis
Research Article
a
Scheme 1. General Reaction Schemes for (A) the Aldol Condensation and (B) the Nitroaldol Condensation
a
37
1
The dinitroalkane side product in the nitroaldol condensation was not observed by H NMR spectroscopy and is not presented in this scheme.
Figure 3. Hammett plot for para-substituted benzaldehydes in the
aldol condensation with acetone at 50 °C. The TOF values were
obtained using the initial rate over the first hour of reaction. The TOF0
value is the TOF obtained when using the reference benzaldehyde (p-
H).
Figure 4. Hammett plot for para-substituted benzaldehydes in the
nitroaldol condensation with nitromethane at 40 °C using σ values.
The TOF values were obtained using the initial rate over the first hour
+
of reaction. The TOF value is the TOF obtained when using the
0
reference benzaldehyde (p-H).
Scheme 2. Proposed Mechanism for the Aldol Condensation of Acetone and Para-Substituted Benzaldehydes on Aminosilica
a
Cooperative Catalysts
a
Generation of the nucleophilic enamine leads to attack on the electrophilic carbonyl group of the aldehyde.
4
64
DOI: 10.1021/acscatal.5b02398
ACS Catal. 2016, 6, 460−468

ACS Catalysis
Research Article
Information for the fit using σ values). As Brown and Okamoto
observed, these modified substituent constants correlate to
cationic species that are generated during the reaction.
However, the Hammett analysis suggests resonance effects
from electron-donating groups in the para position. Previous
reports suggested the formation of imines via attack of the
amines on the aldehydes either when electron-donating groups
were used in the para position of the aldehydes or when the
silanols were capped with trialkylsilyl groups and unable to
41,42
Obtaining a better linear fit of the reactivity data using the σ+
values indicates the formation of a cationic or near-cationic
intermediate or transition state, which suggests that electron-
donating groups in the para position may stabilize these
cationic intermediates via electron delocalization through the
aromatic ring, as will be discussed shortly. A proposed
nitroaldol mechanism is presented in Scheme 3, which begins
3
2,33
participate in cooperative catalysis.
This mechanism
(presented as Scheme S1 in the Supporting Information) is
potentially feasible, as is the mechanism in Scheme 3, though
the subtle effect of the electron-donating group, specifically p-
OMe, is further illustrated in Scheme 4. Here, the carbonyl
group of the aldehyde is stabilized by a surface silanol, which
withdraws electron density from the oxygen atom of the
carbonyl group. To compensate for this effect, the silanol-
stabilized aldehyde has a resonance structure in which the π
electrons of the carbonyl group delocalize to the oxygen atom,
generating an alkoxy ion and leaving a cation at the benzylic
carbon, which is consistent with the improved fit of the
Scheme 3. Proposed Mechanism for the Nitroaldol
Condensation of Nitromethane and Para-Substituted
Benzaldehydes on Aminosilica Cooperative Catalysts
a
+
41,42
Hammett data using the σ values.
When the electron-donating p-OMe group is present, it can
stabilize the benzylic carbocation via electron delocalization
through the aromatic ring, which forms an oxonium ion at the
para position. This through-conjugation or through-resonance
effect of electron-donating groups stabilizes the carbocation
intermediate, which increases the rate of reaction, most likely
by reducing the activation energy. The p-Me substitution likely
does not strongly contribute to a resonance-stabilized form of
the aldehyde, even though it is an electron-donating group. The
resonance structures in Scheme 4 are drawn with the p-OMe
group to illustrate the effect of through-resonance. It is possible
that both proposed mechanisms (Scheme 3 and Scheme S1 in
the Supporting Information) for the nitroaldol condensation
proceed simultaneously for all five benzaldehyde derivatives
examined, though the dominant mechanism likely depends on
the electronic nature of the para substituent.
a
This mechanism postulates the use of two silanols to activate both
24
reactants simultaneously.
with the alkylamine abstracting one of the acidic protons of a
molecule of nitromethane that interacts with a surface silanol.
The aldehyde interacts with a second silanol, which allows for
simultaneous activation of the reactants, and is subjected to
nucleophilic attack at the carbonyl group. The nitroaldol
product is formed, and the cycle is complete. This mechanism
is plausible and is likely the dominant pathway when electron-
withdrawing groups are present in the para position and is
consistent with previous observations of electron-withdrawing
groups in the nitroaldol condensation as reported by Sharma
However, the only data point not to agree with the Hammett
+
analysis using the σ values was the point corresponding to the
p-Br substitution. In spite of the electron-withdrawing proper-
ties of the p-Br group, no plausible resonance structures exist
that can stabilize a carbocation at the benzylic position. If the
substitution is an electron-withdrawing nitro group, plausible
resonance structures exist that would stabilize the same
carbocation via electron delocalization through the aromatic
ring once the carbonyl group hydrogen bonds with the silanol.
These results suggest that para substitutions capable of
3
2
and co-workers.
Scheme 4. Effect of through-Conjugation via Electron Delocalization through the Aromatic Ring of p-OMe-Substituted
a
Benzaldehyde
a
The hydrogen-bonding ability of the silanols stabilizes the aldehyde and withdraws electron density from the carbonyl group, ultimately generating
a benzylic carbocation. The p-OMe group is shown explicitly in the mechanism to illustrate plausible resonance-stabilized carbocation intermediates.
4
65
DOI: 10.1021/acscatal.5b02398
ACS Catal. 2016, 6, 460−468

ACS Catalysis
Research Article
delocalizing electron density throughout the aromatic ring may
accelerate the rate of the nitroaldol condensation, regardless of
whether the substitution is electron-donating or electron-
withdrawing. This suggestion is supported by the reduced
magnitude of the slope of the Hammett plot for the nitroaldol
condensation (0.18) in comparison to that of the aldol
condensation (0.78). It is better to think of the reduced
slope of the Hammett plot in the nitroaldol condensation as the
combined effects of electron-donating groups (improved
stabilization of the reaction intermediate) and electron-
withdrawing groups (increasing the electrophilicity of the
aldehyde) both being capable of increasing the rate of reaction,
albeit for slightly different reasons, which would lead to the
original suggestion that substituents do not have a significant
effect on the kinetics. The reduced magnitude of the slope
could also reflect changes in the dominant mechanism as the
substituent changes, as discussed previously. Furthermore,
these results provide reasonable structures for the likely
reaction intermediates (i.e., the carbocations). The differences
in the Hammett analyses for the aldol and nitroaldol
condensations are in agreement with a previous report
regarding the incorporation of Brønsted and Lewis acid sites
Table 2. Activation Energies, E , and Pre-Exponential
Factors, A, for the Aldol and Nitroaldol Condensations
Using Catalysts with and without Accessible Silanols
a
material
reaction
Ea (kJ/mol)
A (s−1)
6
4
8
8
MCM-C4-0.57-A
aldol
40.1
28.9
51.3
43.5
9.33 × 10
4.03 × 10
8.52 × 10
6.36 × 10
MCM-C4-0.57-A-HMDS
MCM-C4-0.57-A
aldol
nitroaldol
nitroaldol
MCM-C4-0.57-A-HMDS
intermediates as a consequence of the silanols being unavailable
for cooperative interactions. However, it should be noted that
the TOF values for the trimethylsilyl-capped catalysts decreased
in comparison to the catalysts with available silanols and
required higher reaction temperatures to achieve any
appreciable conversion, further emphasizing the importance
of the silanols for cooperative interactions. As an interesting
finding, a significant decrease in the pre-exponential factor
caused the reaction rate to decrease in spite of the decreased
activation energy, providing additional compelling evidence of
the vital role that cooperative acid−base interactions serve in
the aldol condensation. The pre-exponential factor depends on
the total number of molecular collisions occurring over the
course of the reaction, meaning that as the value of the pre-
exponential factor increases, it is more likely that reactant
molecules, most likely including cocatalytic amines and silanols,
are colliding with one another in productive ways with greater
30
into silica materials by heteroatom substitution. Modest
increases in the number of Lewis acid sites were found to
increase the catalytic activity for the nitroaldol condensation,
30
but not for the aldol condensation. This effect is consistent
with the proposed structures in Scheme 4 because the acid
interactions with the carbonyl group of the aldehyde assist the
aldehyde with entering the catalytic cycle and help stabilize the
resulting carbocation. The acid functionality of the surface, be it
through native silanols, heteroatom substitutions, or other acid
sources/sites, could help withdraw the electron density from
the carbonyl group, making the carbon at the benzylic position
more electrophilic, facilitating both through-resonance from
electron-donating groups at the para position and easier
nucleophilic attack at the carbocation. Conversely, stabilizing
the aldehyde in the aldol condensation would promote the
nucleophilic attack of the aldehyde by the amine to form an
imine (a secondary aldimine), rendering the amine less active
for catalysis because the acetone would be unable to react with
43
frequency. After the silanols are capped, the reaction has a
similar or even lower energy transition state, but the likelihood
of reaching that state by productive collisions is far lower due to
the lack of available silanols, which were previously abundantly
present. In this regard, capping the silanols significantly
decreased the pre-exponential factor because the acid sites
were unable to promote the activation of acetone and likely led
to the amines attacking the aldehyde, forming unproductive
imines (secondary aldimines) from the aldehydes, decreasing
the number of available amines in the catalytic cycle. These
results are also consistent with previous reports in which
stronger acid sites decreased the catalytic activity of amino-
silicates in the aldol condensation, in the sense that substituting
stronger acid sites for the inherently weaker silanols decreased
the observed rates of reaction, but some acids were necessary to
31
the resulting imine. In this case, the aldimine cannot form a
nucleophilic enamine because there are no α protons available
on the aldimine, which bears an aryl substituent, resulting in the
10,11,22,27,30
achieve appreciable conversion.
Another possible
31
aldimine being a catalytically inactive off-cycle species. Unlike
deprotonated nitromethane, which can react with the aldimine
in the nitroaldol condensation to form the product as shown in
Scheme S1 in the Supporting Information, acetone cannot be
activated in such a way to make it a better nucleophile; thus, if
the amines preferentially attack the acid-stabilized aldehyde in
the aldol condensation, the catalytic activity should decrease.
Arrhenius Analysis of Cooperative Catalysts. Arrhenius
analysis of an aminosilica catalyst with available silanols (MCM-
C4-0.57-A) and with trimethylsilyl-capped silanols (MCM-C4-
interpretation of the reduced activation energy is that the
reduced value reflects the direct nucleophilic attack of the
acetone by the amines without the assistance of the silanols. If a
silanol is present, it will hydrogen bond with the acetone,
meaning that the energy of the hydrogen bond must be
accounted for in the activation of the acetone by the amine.
Without available silanols, this hydrogen bond does not need to
be disrupted, which could explain why the activation energy
decreased for the aldol condensation after capping the silanols.
This direct acetone activation pathway is consistent with the
decreased pre-exponential factor, because the number of direct
molecular interactions of the amines with the acetone
molecules likely would be reduced dramatically without silanols
available for cooperative interactions.
The Arrhenius analysis for the nitroaldol conversion further
revealed differences between the two condensation reactions.
The obtained activation energies and pre-exponential factors
were only slightly decreased for the aminosilica catalysts with
capped silanols in comparison to the catalysts with available
silanols, indicating that the loss of the acid functionality of the
0
.57-A-HMDS) revealed very useful mechanistic insight into
the aldol and nitroaldol condensations from the obtained
activation energies and pre-exponential factors reported in
Table 2. All Arrhenius plots are presented in Figures S2−S5 in
the Supporting Information, and the TOFs are reported in
Tables S1 and S2 in the Supporting Information. For the aldol
condensation, the activation energy and the pre-exponential
factor decreased once the silanols were capped with
trimethylsilyl groups. The observed decrease in activation
energy could be due to the presence of different reaction
4
66
DOI: 10.1021/acscatal.5b02398
ACS Catal. 2016, 6, 460−468

ACS Catalysis
Research Article
surface was not as severe as in the case of the aldol
condensation. At first, this notion would seem to be
counterintuitive on the basis of the evidence that increased
numbers of acid sites have been shown to increase the activity
withdrawing electron density from the carbonyl group via the
aromatic ring. Conversely, in the nitroaldol condensation, the
p-OMe electron-donating group provided additional stability to
the intermediate carbocation generated at the benzylic position
via through-resonance in the form of electron delocalization
through the aromatic ring. Arrhenius analysis of the two
condensations revealed a stronger dependence of the pre-
exponential factor for the aldol condensation, owing to the
difficulty of activating the acetone molecules in the presence of
capped silanols on the silica surface (i.e., loss of acid character),
suggesting that the amines and silanols were both necessary to
promote molecular collisions necessary for cooperative
catalysis. However, capping the silanols with trimethylsilyl
groups for all reactions required elevated temperatures to
achieve appreciable conversion and reduced the observed TOF
values in comparison to materials with available silanols, further
emphasizing the importance of the silanols as weak acid
partners in cooperative catalysis.
30
in the nitroaldol condensation, but other observations may
lead to a plausible explanation. First, the nitroaldol con-
densation may readily proceed via a base-catalyzed mechanism
in which no acid components are necessary to complete the
3
3−35
catalytic cycle.
This proposed mechanism involves the
formation of an imine intermediate via condensation of the
aldehyde with an amine with subsequent attack of the
deprotonated nitromethane on the imine to form the nitroaldol
33−35
product (see Scheme S1 in the Supporting Information).
In contrast, the aldol condensation requires activation of the
acetone and subsequent attack of the enamine on the carbonyl
of the aldehyde. If the amines dehydrate the aldehyde first, the
resulting imine most likely would be unable to react with the
acetone molecules present in solution, meaning that a base-only
mechanism likely would not effectively promote the aldol
condensation using aminosilica catalysts with capped silanols
ASSOCIATED CONTENT
■
31
(
i.e., without acid sites). This mechanism would also explain
*
S
Supporting Information
why capping the silanols was severely detrimental to the aldol
condensation yet not as disadvantageous to the activity of the
nitroaldol condensation. Most likely, the array of proposed
mechanisms of the nitroaldol condensation operate simulta-
neously and not all of them require silanols or they can use
other acid sources; thus, when the silanols are capped, there are
alternative pathways for the reaction to proceed in the absence
of silanols.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acscatal.5b02398.
Hammett plot of the nitroaldol reactivity data using σ
values, the base-only nitroaldol mechanism that operates
via imine intermediate, Arrhenius plots and TOFs, and a
TEM image of the MCM-41 (PDF)
Previous work demonstrated that changing the alkylamine
linker length and the pore diameter of the silica (i.e., the
curvature of the surface) had a stronger effect on the observed
TOFs for the nitroaldol condensation than for the aldol
AUTHOR INFORMATION
■
*
Corresponding Authors
E-mail for E.G.M.: eric.moschetta@chbe.gatech.edu.
*E-mail for C.W.J.: cjones@chbe.gatech.edu.
2
4
condensation. With the results of the present work, these
linker length and surface curvature effects might be viewed as a
likely consequence of allowing each available alkylamine to
interact with a larger total number of available silanols, which
would have a more pronounced effect on the nitroaldol
condensation (two possible silanol-assisted pathways) in
comparison to the aldol condensation (one silanol-assisted
pathway). As with the aldol condensation, it should be noted
that the TOF values decreased using the catalyst with the
capped silanols in comparison to the catalyst without capped
silanols and elevated temperatures were required to achieve
appreciable conversion. The larger decrease in the pre-
exponential factor for the aldol condensation when the silanols
were unavailable for cooperative interactions suggests that the
aldol condensation is more dependent on the amines and the
silanols being fully accessible to facilitate cooperative molecular
collisions between the reactants in comparison to the nitroaldol
condensation.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the Department of Energy Office
of Basic Energy Sciences through Catalysis Contract No.
DEFG02-03ER15459. The authors thank Dr. Li-Chen Lee and
Ms. Caroline Hoyt for helpful discussions, Dr. Li-Chen Lee for
performing the NMR experiments, and Dr. Simon Pang for
taking the TEM image. The authors thank a reviewer for
particularly insightful comments.
REFERENCES
■
(
(
1) Trost, B. M. Angew. Chem., Int. Ed. Engl. 1995, 34, 259−281.
2) Ritleng, V.; Sirlin, C.; Pfeffer, M. Chem. Rev. 2002, 102, 1731−
770.
1
(
3) Colby, D. A.; Bergman, R. G.; Ellman, J. A. Chem. Rev. 2010, 110,
24−655.
(4) Tsubogo, T.; Ishiwata, T.; Kobayashi, S. Angew. Chem., Int. Ed.
2013, 52, 6590−6604.
5) Finelli, F. G.; Miranda, L. S. M.; de Souza, R. O. M. A. Chem.
Commun. 2015, 51, 3708−3722.
6) Vural Gursel, I.; Noel, T.; Wang, Q.; Hessel, V. Green Chem.
015, 17, 2012−2026.
7) Margelefsky, E. L.; Zeidan, R. K.; Davis, M. E. Chem. Soc. Rev.
008, 37, 1118−1126.
8) Bornscheuer, U. T.; Huisman, G. W.; Kazlauskas, R. J.; Lutz, S.;
6
CONCLUSIONS
■
The kinetics and mechanisms of cooperatively catalyzed aldol
and nitroaldol condensations using amine-functionalized
MCM-41 mesoporous silicas as bifunctional acid−base catalysts
were examined using Hammett and Arrhenius analyses. From
the Hammett analysis of the aldol condensation, electron-
withdrawing para substituents on benzaldehyde were found to
accelerate of the rate of the reaction. This observation was
consistent with the proposed mechanism, in which the carbonyl
of the aldehyde is subjected to nucleophilic attack by an
enamine, which would be accelerated by para substituents
(
(
2
(
2
̈
̈
(
Moore, J. C.; Robins, K. Nature 2012, 485, 185−194.
(9) Warshel, A.; Sharma, P. K.; Kato, M.; Xiang, Y.; Liu, H.; Olsson,
M. H. M. Chem. Rev. 2006, 106, 3210−3235.
4
67
DOI: 10.1021/acscatal.5b02398
ACS Catal. 2016, 6, 460−468

ACS Catalysis
Research Article
(
10) Zeidan, R. K.; Hwang, S. J.; Davis, M. E. Angew. Chem., Int. Ed.
2
(
(
006, 45, 6332−6335.
11) Zeidan, R. K.; Davis, M. E. J. Catal. 2007, 247, 379−382.
12) Shylesh, S.; Wagener, A.; Seifert, A.; Ernst, S.; Thiel, W. R.
ChemCatChem 2010, 2, 1231−1234.
(13) Shylesh, S.; Thiel, W. R. ChemCatChem 2011, 3, 278−287.
(14) Shylesh, S.; Hanna, D.; Gomes, J.; Canlas, C. G.; Head-Gordon,
M.; Bell, A. T. ChemSusChem 2015, 8, 466−472.
15) Hanna, D. G.; Shylesh, S.; Li, Y.-P.; Krishna, S.; Head-Gordon,
M.; Bell, A. T. ACS Catal. 2014, 4, 2908−2916.
16) Shylesh, S.; Hanna, D.; Gomes, J.; Krishna, S.; Canlas, C. G.;
Head-Gordon, M.; Bell, A. T. ChemCatChem 2014, 6, 1283−1290.
17) Shylesh, S.; Sreekumar, S.; Gomes, J.; Grippo, A.; Arab, G. E.;
Head-Gordon, M.; Toste, F. D.; Bell, A. T. Angew. Chem., Int. Ed.
015, 54, 4673−4677.
18) Gianotti, E.; Diaz, U.; Velty, A.; Corma, A. Catal. Sci. Technol.
013, 3, 2677−2688.
19) Díaz, U.; Brunel, D.; Corma, A. Chem. Soc. Rev. 2013, 42, 4083−
097.
20) Bass, J. D.; Solovyov, A.; Pascall, A. J.; Katz, A. J. Am. Chem. Soc.
(
(
(
2
(
2
(
4
(
2
(
(
006, 128, 3737−3747.
21) Hruby, S. L.; Shanks, B. H. J. Catal. 2009, 263, 181−188.
22) Brunelli, N. A.; Venkatasubbaiah, K.; Jones, C. W. Chem. Mater.
2
(
012, 24, 2433−2442.
23) Brunelli, N. A.; Didas, S. A.; Venkatasubbaiah, K.; Jones, C. W. J.
Am. Chem. Soc. 2012, 134, 13950−13953.
(24) Brunelli, N. A.; Jones, C. W. J. Catal. 2013, 308, 60−72.
(25) Lauwaert, J.; De Canck, E.; Esquivel, D.; Thybaut, J. W.; Van
Der Voort, P.; Marin, G. B. ChemCatChem 2014, 6, 255−264.
26) Lauwaert, J.; De Canck, E.; Esquivel, D.; Van Der Voort, P.;
Thybaut, J. W.; Marin, G. B. Catal. Today 2015, 246, 35−45.
27) Lauwaert, J.; Moschetta, E. G.; Van Der Voort, P.; Thybaut, J.
W.; Jones, C. W.; Marin, G. B. J. Catal. 2015, 325, 19−25.
28) Balakrishnan, M.; Sacia, E. R.; Bell, A. T. ChemSusChem 2014, 7,
078−1085.
29) Sharma, K. K.; Buckley, R. P.; Asefa, T. Langmuir 2008, 24,
4306−14320.
30) Moschetta, E. G.; Brunelli, N. A.; Jones, C. W. Appl. Catal., A
015, 504, 429−439.
31) Kandel, K.; Althaus, S. M.; Peeraphatdit, C.; Kobayashi, T.;
Trewyn, B. G.; Pruski, M.; Slowing, I. I. J. Catal. 2012, 291, 63−68.
32) Sharma, K. K.; Biradar, A. V.; Asefa, T. ChemCatChem 2010, 2,
1−66.
33) Bass, J. D.; Solovyov, A.; Pascall, A. J.; Katz, A. J. Am. Chem. Soc.
006, 128, 3737−3747.
34) Demicheli, G.; Maggi, R.; Mazzacani, A.; Righi, P.; Organica, C.;
Mangini, A. Tetrahedron Lett. 2001, 42, 2401−2403.
35) Zorn, D.; Lin, V. S. Y.; Pruski, M.; Gordon, M. S. J. Phys. Chem.
A 2008, 112, 10635−10649.
36) Lukens, W. W.; Schmidt-Winkel, P.; Zhao, D.; Feng, J.; Stucky,
G. D. Langmuir 1999, 15, 5403−5409.
37) Motokura, K.; Tada, M.; Iwasawa, Y. Angew. Chem., Int. Ed.
(
(
(
1
(
1
(
2
(
(
6
(
2
(
(
(
(
2
(
(
008, 47, 9230−9235.
38) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165−195.
39) Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck,
J. S. Nature 1992, 359, 710−712.
40) Kruk, M.; Jaroniec, M.; Kim, J. M.; Ryoo, R. Langmuir 1999, 15,
279−5284.
41) Brown, H. C.; Okamoto, Y. J. Am. Chem. Soc. 1957, 79, 1913−
917.
42) Brown, H. C.; Okamoto, Y. J. Am. Chem. Soc. 1958, 80, 4979−
987.
43) Moore, J. W.; Pearson, R. G. In Kinetics and Mechanism; Wiley:
New York, 1981; pp 116−121.
(
5
(
1
(
4
(
4
68
DOI: 10.1021/acscatal.5b02398
ACS Catal. 2016, 6, 460−468