Effects of amine structure and base strength on acid-base cooperative aldol condensation

Article abstract of DOI:10.1016/j.cattod.2014.08.007

Aminated silica materials are known to efficiently catalyse aldol condensations, especially when silanol groups are neighbouring the amine function. The effect of the amine structure and base strength has been analysed experimentally and by kinetic modelling using commercially available precursors to graft primary, secondary and tertiary amines on the silica surface. While primary amines are arranged in a clustered manner on the catalyst surface, secondary amines are arranged randomly which results in a higher percentage of promoted amines in the low silanol-to-amine ratio range. An enamine compound formed by the reaction between the amine active site and acetone has been identified as the key intermediate to explain the experimental observations. In the case of a primary amine this enamine intermediate can form an inhibiting imine with which it is in equilibrium. As a secondary amine has only one hydrogen atom bonded to the nitrogen atom, the inhibiting imine cannot be formed, resulting in a comparatively higher concentration of reactive enamines on the catalyst surface. In case of a tertiary amine the formation of the reactive intermediate is impossible due to the absence of any hydrogen atom bonded to the nitrogen atom. The activation entropies of all reaction steps occurring on the amine sites, as obtained by regression, could be correlated to the deprotonation entropies of the amine sites. As the deprotonation enthalpy does not account for steric effects, no such correlation could be found between the activation energies of these reaction steps and the deprotonation enthalpies of the amine sites.

Full text of DOI:10.1016/j.cattod.2014.08.007

G Model
CATTOD-9214; No. of Pages11
ARTICLE IN PRESS
Catalysis Today xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Effects of amine structure and base strength on acid–base cooperative
aldol condensation
Jeroen Lauwaerta, Els De Canckb, Dolores Esquivelb, Pascal Van Der Voortb,
Joris W. Thybauta,∗, Guy B. Marina
a Ghent University, Department of Chemical Technology, Laboratory for Chemical Technology, Technologiepark 914, 9052 Ghent, Belgium
b Ghent University, Department of Inorganic and Physical Chemistry, Center for Ordered Materials, Organometallics & Catalysis, Krijgslaan 281 (S3), 9000
Ghent, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online xxx
Aminated silica materials are known to efficiently catalyse aldol condensations, especially when silanol
groups are neighbouring the amine function. The effect of the amine structure and base strength has
been analysed experimentally and by kinetic modelling using commercially available precursors to graft
primary, secondary and tertiary amines on the silica surface. While primary amines are arranged in a
clustered manner on the catalyst surface, secondary amines are arranged randomly which results in a
higher percentage of promoted amines in the low silanol-to-amine ratio range. An enamine compound
formed by the reaction between the amine active site and acetone has been identified as the key interme-
diate to explain the experimental observations. In the case of a primary amine this enamine intermediate
can form an inhibiting imine with which it is in equilibrium. As a secondary amine has only one hydrogen
atom bonded to the nitrogen atom, the inhibiting imine cannot be formed, resulting in a comparatively
higher concentration of reactive enamines on the catalyst surface. In case of a tertiary amine the forma-
tion of the reactive intermediate is impossible due to the absence of any hydrogen atom bonded to the
nitrogen atom. The activation entropies of all reaction steps occurring on the amine sites, as obtained by
regression, could be correlated to the deprotonation entropies of the amine sites. As the deprotonation
enthalpy does not account for steric effects, no such correlation could be found between the activation
energies of these reaction steps and the deprotonation enthalpies of the amine sites.
Keywords:
Aldol reaction
Amines
Silanols
Kinetic modelling
Acid–base
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
Aldol condensations are important reactions, typically
employed in the pharmaceutical industry and fine chemicals
There are some inherent disadvantages to homogeneous catal-
ysis such as (i) the necessity of energy intensive separation steps,
(ii) a short catalyst life time and (iii) the low reusability, (iv)
important waste streams and (v) intensive equipment corrosion
[1,2]. Heterogeneous catalysis may help to overcome many of
these disadvantages. Therefore, it is important to develop new
and optimize existing heterogeneous catalysts that can, poten-
tially, replace homogeneous ones. Since about a decade, so-called
high-throughput technologies have emerged in order to speed up
catalyst development and optimization. However, the discovery
rate may be further enhanced by exploiting more rational feed-
back obtained by means of kinetic modelling in the catalyst design
cycle [3–5].
production, to create new C C bonds and, hence, heavier and
more complex molecules [6–9]. A bright future has also already
been forecasted for aldol condensation reactions in the transition
from a fossil resources based towards a more sustainable society,
e.g., in the valorization of glycerol, which is a byproduct in the
production of biodiesel, or the conversion of furan compounds such
as 2-furaldehyde (furfural) and 5-(hydroxymethyl)furfural (HMF),
which are obtained by dehydration of sugars, into hydrocarbon
fuels [10–15]. At present aldol condensations are industrially
mostly catalysed by strong, homogeneous base catalysts such
as KOH, Ca(OH)2, NaOH or Na2CO3. Recent research focuses on
hydrotalcite catalysts and other layered double hydroxides [16–21]
or functionalized silica materials [22–35] as possible alternatives.
Silica materials on which amine containing silanes are grafted
by means of a stirring or a reflux procedure have been found to
perform well [24–32]. Moreover, many authors agree that the
incorporation of weak acid sites, such as silanol groups, next to the
∗
Corresponding author. Tel.: +32 09 331 17 57.
E-mail address: Joris.Thybaut@UGent.be (J.W. Thybaut).
http://dx.doi.org/10.1016/j.cattod.2014.08.007
0920-5861/© 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: J. Lauwaert, et al., Effects of amine structure and base strength on acid–base cooperative aldol
condensation, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.08.007

G Model
CATTOD-9214; No. of Pages11
ARTICLE IN PRESS
2
J. Lauwaert et al. / Catalysis Today xxx (2014) xxx–xxx
Fig. 1. Representation of the catalyst surfaces.
base amine sites has a enhancing effect on the catalytic activity
of these amines in aldol condensations and other important C
atmosphere at 110 ◦C for 24 h. The solid is recovered by filtration
and is, subsequently, diluted in chloroform (>99.8%, Roth) and
stirred for 3 h in order to remove all the remaining toluene and
precursor from the pores. After drying in vacuo at room temper-
ature for 24 h the cooperative acid–base catalysts (samples AB)
were ready for kinetic testing.
C
coupling reactions [22–30,33–38]. Recently it has been demon-
strated that an increasing acid strength of the promoting site leads
to a decreasing activity which could be explained by a pronounced
shift in the equilibrium from the free acid and free base towards
the resulting neutralized ion pair [25,26,32].
In this work, a kinetic model is developed for the acid–base
cooperative catalysed aldol condensation reaction. The focus is on
the effect of the structure and base strength of the amine func-
tional groups on the aldol condensation kinetics. A catalyst library
comprising both base and acid–base catalysts was synthesized by
functionalization of a commercial, mesoporous silica, Silicagel 60,
with different commercially available primary, secondary and ter-
tiary amine containing silanes. The effect of the reaction conditions
as well as the structure and base strength of the active site on the
catalyst activity is investigated by means of both an experimental
kinetic study and kinetic modelling.
2.1.2. Endcapping of the silanol groups in samples AB with
1,1,1,3,3,3-hexamethyldisilazane
About half of each sample of the cooperative acid–base
catalysts (samples AB) was subsequently treated with 1,1,1,3,3,3-
hexamethyldisilazane (98%, HMDS, ABCR) in order to prepare the
base catalysts (samples B). 2.5 g of the dried cooperative catalyst
is added to a flask and completely covered with 30 ml of HMDS.
The mixture is vigorously stirred at room temperature for 3 h and
subsequently, the mixture is filtered and thoroughly washed with
chloroform for 3 h using the same method as during the grafting of
the amine containing silanes. Finally, the solids are dried in vacuo
at room temperature for 24 h. HMDS quantitatively reacts with the
surface silanols of silicagel if the silica is calcined at a tempera-
ture above 400 ◦C [42,43]. Hence, this HMDS treatment endcaps all
remaining silanols on the surface by replacing it with a trimethylsi-
lyl function. The synthesis procedure finally led to 10 different
catalysts as represented in Fig. 1.
2. Procedures
2.1. Preparation of catalysts
2.1.1. Grafting of amines on Silicagel 60
Different catalysts were prepared using five commer-
cially available amine containing silanes, (3-aminopropyl)
triethoxysilane (98%, APTES, ABCR), N-methylaminopropyl-
trimethoxysilane (MAPTMS, ABCR), N-cyclohexylaminopropyl-
trimethoxysilane (CAPTMS, ABCR), N-phenylaminopropyl-
trimethoxysilane (95% PAPTMS, ABCR) and (N,N-diethyl-
3-aminopropyl)trimethoxysilane (DEAPTMS, ABCR). These
2.2. Catalyst characterization
Nitrogen adsorption–desorption measurements are carried out
at 77 K using a Belsorp Mini II gas analyser. Samples are degassed at
120 ◦C for 17 h prior to measurement. The specific surface area and
pore volume are determined using the Brunauer–Emmett–Teller
(BET) method. The average pore size of the silica is obtained using
the Barrett–Joyner–Halenda (BJH) method.
The concentrations of amine and silanol groups are determined
using elemental (CHNS) analysis. These experiments are performed
on a Thermo Flash 2000 elemental analyser using V2O5 as catalyst.
Sample cups are tightened and the analysis waiting time is kept
as short as possible to avoid, or at least minimize, readsorption of
water, impacting on the weight and elemental composition of the
samples.
The presence of amine groups after grafting and silanol removal
through HMDS treatment is demonstrated by means of Diffuse
Reflectance Infrared Fourier Transform (DRIFT) spectroscopy. These
measurements are performed on a Nicolet 6700 of Thermo Scien-
tific with a nitrogen cooled MCT-A detector. The DRIFT spectra are
obtained using a Graseby Specac diffuse reactant cell, operating
in vacuo at 140 ◦C.
aminosilanes were grafted on
a mesoporous silica, Silicagel
60 (Grade 7734, Sigma–Aldrich). First, the silica is calcined at
700 ◦C for 6 h with a heating rate of 2 ◦C/min. After the calcination
the material is slowly cooled to a temperature of about 150 ◦C.
5 g of pretreated material is diluted in 30 ml of toluene (Extra
Dry, Acros) while still hot to avoid rehydration of the samples.
The aminosilane is subsequently added. The added volume of the
amine precursor is varied between several batches in order to
silylate the silica with different amounts of amine groups. The
molar ratio of added precursor to free silanols is varied in the range
of 0.1–5. The necessary volume of precursor is calculated assuming
the number of free silanols is equal to 1.1 OH/nm2 [39]. However,
caution is required as the silica, even after a treatment at 700 ◦C
can still readsorb water, and variations up to 25% are possible
when comparing different kinds of silica [40,41]. Therefore, it is
very important to individually characterize each catalyst batch.
After adding the precursor, the mixture is refluxed under an argon
Please cite this article in press as: J. Lauwaert, et al., Effects of amine structure and base strength on acid–base cooperative aldol
condensation, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.08.007

G Model
CATTOD-9214; No. of Pages11
ARTICLE IN PRESS
J. Lauwaert et al. / Catalysis Today xxx (2014) xxx–xxx
3
Scheme 1. Aldol condensation of acetone and 4-nitrobenzaldehye towards 4-hydroxy-4-(4-nitrophenyl)butane-2-one and 4-(4-nitrophenyl)-3-buten-2-one, the aldol and
ketone, respectively.
2.3. Catalyst performance testing
and the amine site which yields an inhibiting imine [22,23]. In
order to investigate the effect of the catalyst properties on the aldol
condensation performance, all catalysts were subject to identical
operating conditions, i.e., 318 K with 0.79 mol acetone, 3.31 mmol
4-nitrobenzaldehyde, 0.39 mol n-hexane (solvent) and 1.44 mmol
methyl 4-nitrobenzoate (IS). Note that, due to the different amine
loading on the different catalysts, the added catalyst mass has to be
altered in order to keep the concentration of amine groups in the
reactor identical in all experiments.
Catalysts performance has been assessed via the aldol con-
densation of acetone (99.6%, Acros) and 4-nitrobenzaldehyde
(99%, Acros) towards the primary aldol product 4-hydroxy-4-
(4-nitrophenyl)butan-2-one and the secondary ketone product
4-(4-nitrophenyl)-3-buten-2-one, the latter being formed after
elimination of water from the aldol (Scheme 1). The experiments
with the APTES functionalized catalysts are carried out in a Parr
4560 mini reactor, which is a batch type reactor with a volume
of 300 ml. The temperature in the reactor is maintained using a
PID-controller (CAL 9500P controller) and a thermocouple inside
the reactor vessel. The PID-controller is connected to the reactor
heating jacket and an additional cooling unit (DLK 402 circulating
cooler). The reaction mixture is stirred with a mechanical stirrer
rotating at 500 rpm. The experiments with the other catalysts are
carried out in a three-necked glass flask of 180 ml equipped with
a reflux condenser, a thermocouple and a sampling port, which is
also a batch type reactor. The temperature in this reactor is main-
tained using a Lauda Proline RP845 PID controller and the reaction
mixture is stirred with a magnetic stirrer rotating at 500 rpm. Some
experiments were carried out in both reactors to confirm that the
deviations on the kinetic data due to the differences in the setup
are negligible.
2.4. Thermodynamic non-ideality of the liquid mixture
As the experiments are performed in a liquid phase comprising
components with different polarities, it is important for the model
to explicitly account for the non-ideality of the liquid phase via so-
called activity coefficients, ꢀi [45–47]. The Universal Quasichemical
(UNIQUAC) method and its extension, the UNIQUAC Functional-
Group Activity Coefficients (UNIFAC) method are widely accepted
methods to calculate activity coefficients and other thermophysical
data of compounds in liquid multi-component mixtures without
explicit use of experimental data. Both methods make use of binary
interaction parameters to describe the interactions between two
molecules or two functional groups in the mixture. More detailed
information about the mathematics of these methods can be found
in the Supporting Information.
First, the reactor is loaded with the catalyst, n-hexane (Extra
Pure, Acros) as solvent and methyl 4-nitrobenzoate as internal stan-
dard (IS, >99%, Aldrich). Then the mixture is heated to the desired
reaction temperature. Acetone is separately heated to the reac-
tion temperature and used to dissolve 4-nitrobenzaldehyde before
injection in the reactor. The time of injection is taken as the start of
the reaction (t = 0). The reaction is monitored for 200 min by taking
a sample (about 0.5 ml) of the reaction mixture every 20 min. For
each experiment, the total decrease of reaction volume due to sam-
pling is less than 5% and the effect on the kinetic data is considered
to be negligible. The samples are analysed using a reversed-phase
high-performance liquid chromatograph (RP-HPLC), from Agilent
(1100 series). The HPLC is operated at a column temperature of
30 ◦C using a gradient method with water (0.1% trifluoroacetic
acid, Acros) and acetonitrile (HPLC grade, Acros) as solvents. In this
gradient method the volumetric percentage of acetonitrile is var-
ied from 30% to 62% over a period of 7 min. The components are
identified using a UV-detector with a variable wavelength. Quan-
tification of the different components in the reaction mixture is
performed by relating the peak surface areas to the amount of inter-
nal standard, methyl 4-nitrobenzoate. In this work, the activity of
the catalysts is described by the turnover frequency (TOF), which
could be determined from the slope of the initial linear part of
the conversion of 4-nitrobenzaldehyde as a function of time, the
concentration of amine active sites and the initial concentration of
4-nitrobenzaldehyde [44]. The concentration of the silanol groups
is not considered in the calculation of the TOF because the silanol
groups only act as promoters and not as actual, catalytically active
sites.
A large number of the interaction parameters used in these mod-
els can be found in literature [48], however, even to date some of the
parameters remain unknown. The thermodynamic non-ideality of
the liquid toluene–precursor mixture used in the synthesis of the
catalysts could be described using the UNIFAC method but some
interaction parameters necessary to describe the non-ideality of the
reaction mixture are unknown. In addition, this model has another
limitation, namely the fundamental assumption that a contribu-
tion made by one group is independent of that made by another
group. This assumption is only valid when neighbouring effects are
absent, which is unlikely when, e.g., a carbon–carbon double bond
is conjugated with an aromatic ring.
An alternative to the above described methods is the Conductor-
Like Screening Model for Real Solvents (COSMO-RS) [49–52] which
is a statistical thermodynamics treatment of the solute–solvent
interactions. The method calculates the thermodynamic data from
molecular surface polarity distributions, which results from quan-
tum chemical calculations of the individual compounds in the
mixture performed using TURBOMOLE. In these quantum chem-
ical calculations the molecules are considered to be in a conductor
environment in which the solute molecule induces a polarization
Table 1
Range of experimental conditions.
Variable
Purpose
Range
Amine groups
Silanol groups
4-Nitrobenzaldehyde
Acetone
Methyl 4-nitrobenzoate
n-Hexane
Active site
Active site promotion
Reactant 1
Reactant 2
Internal standard
Solvent
1 mmol/l
0–2.6 mmol/l
15–177 mmol/l
1.4–7.0 mol/l
13 mmol/l
The range of experimental conditions used to investigate the
effect of the reaction conditions on the catalysts’ activity is given
in Table 1. An acetone excess is used in all experiments in order
to suppress the direct interaction between 4-nitrobenzaldehyde
3.7–6.4 mol/l
Temperature
15–55 ◦
C
Please cite this article in press as: J. Lauwaert, et al., Effects of amine structure and base strength on acid–base cooperative aldol
condensation, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.08.007

G Model
CATTOD-9214; No. of Pages11
ARTICLE IN PRESS
J. Lauwaert et al. / Catalysis Today xxx (2014) xxx–xxx
Table 2
4
charge density on the interface between the molecule and the con-
ductor. These charges act back to the solute and generate a more
polarized electron density than in vacuum. This polarization charge
density is, subsequently, used by the COSMOthermX programme
for the calculation of the interaction energy of the most impor-
tant molecular interaction modes, i.e., electrostatics and hydrogen
bonds.
In this work, the activity coefficients of all components in the
reaction mixture are, in the first instance, calculated using the
COSMO-RS method, in order to overcome the limitations of the UNI-
QUAC and UNIFAC models. In these calculations the liquid phase
composition and temperature are varied in the range of the exper-
imental conditions in order to calculate a reference set of activity
coefficients This set is subsequently used to estimate the binary
interaction parameters in the UNIQUAC method, see Table 5. Due
to the presence of the solvent, i.e., n-hexane, and the use of an ace-
tone excess in the aldol condensation experiments, a component
is only assumed to interact with n-hexane or acetone and all other
binary interaction parameters are neglected. Finally, the UNIQUAC
model with the estimated binary interaction parameters is imple-
mented in the kinetic model to account for the thermodynamic
non-ideality of the reaction mixture.
Concentration of active sites determined from CHNS elemental analysis.
Catalyst
Aminea
Silanolb
Silanol/aminec
[mmol g−1
]
[mmol g−1
]
APTES 1
APTES 2
APTES 3
APTES 4
APTES 5
APTES 6
APTES 7
APTES 8
0.29
0.32
0.33
0.35
0.40
0.49
0.59
0.65
0.67
0.69
0.67
0.61
0.56
0.50
0.45
0.27
2.33 0.13
2.15 0.11
2.07 0.11
1.73 0.09
1.39 0.07
1.03 0.04
0.77 0.04
0.42 0.03
MAPTMS 1
MAPTMS 2
MAPTMS 3
MAPTMS 4
MAPTMS 5
MAPTMS 6
MAPTMS 7
0.34
0.35
0.41
0.47
0.61
0.70
0.80
0.67
0.67
0.66
0.51
0.35
0.31
0.13
1.96 0.10
1.90 0.10
1.60 0.08
1.09 0.05
0.57 0.03
0.44 0.02
0.17 0.02
CAPTMS 1
CAPTMS 2
CAPTMS 3
CAPTMS 4
0.30
0.43
0.56
0.74
0.67
0.60
0.40
0.25
2.20 0.14
1.40 0.06
0.72 0.04
0.35 0.03
PAPTMS 1
PAPTMS 2
0.29
0.58
0.66
0.37
2.23 0.16
0.63 0.04
DEAPTMS 1
DEAPTMS 2
DEAPTMS 3
DEAPTMS 4
0.33
0.52
0.55
0.60
0.66
0.48
0.42
0.45
2.03 0.12
0.93 0.04
0.77 0.04
0.76 0.04
2.5. Joback method
The method of Joback [53] is a relatively simple group con-
tribution method which can be used to estimate the enthalpy of
formation, the enthalpy of vapourization, the Gibbs free energy of
formation and the heat capacity of individual molecules. In this
work, this method is used to estimate the equilibrium coefficient
between the enamine intermediate and the inhibiting imine as well
as the difference between the activation energies of the forward
and backward reaction steps between the aldol and the ketone
products, see Section 4.3.
a
The amine concentration on the cooperative acid–base (type AB) and the corre-
sponding base (type B) catalysts is identical. The error on this measurement is for
all samples 0.01 mmol g−1
.
b
The reported silanol concentration is only valid for cooperative acid–base cata-
lysts (type AB). The error on this measurement amounts to 0.01 mmol g−1 for all
samples.
c
The ratio of silanols to amines given in the table is only valid for cooperative
acid–base catalysts (type AB).
of 497 m2/g, a total pore volume of 0.69 cm3/g and an average pore
diameter of 5.6 nm.
2.6. Reactor model
After grafting the precursor on the silica material the pres-
ence of amine groups is demonstrated via Diffuse Reflectance
Infrared Fourier Transform (DRIFT) measurements, see Fig. 2. The
unfunctionalized pretreated silica exhibits a narrow vibration band
at 3745 cm−1 which is characteristic for the free silanol groups
(Si OH), see Fig. 2(1). After treatment with the aminosilane sev-
eral new bands appear in the range of 2800–3000 cm−1, which
can be attributed to the C H stretch of the grafted functional
groups. The spectrum of the silica material functionalized with
phenylaminopropane active sites (Fig. 2(5)), also shows two addi-
Since the experiments have been performed in a batch type
reactor, which is considered to be spatially uniform in both com-
position and temperature, the mass balance for species i is given
by Eq. (1). With t representing the time and Ci and Ri represent-
ing, respectively, the concentration and the net rate of formation
of component i:
dCi
= Ri
(1)
dt
tional bands at 1505 and 1603 cm−1 due to stretching of the C
C
2.7. Regression analysis
bonds of the aromatic ring. The primary amine group in amino-
propane (Fig. 2(2)) exhibits two small but significant bands at
3314 and 3380 cm−1. The secondary amines methylaminopropane
(Fig. 2(3)) and cyclohexylaminopropane (Fig. 2(4)) gives only one
The model parameters were estimated based on Bayes’ theorem
using Athena Visual Studio [54]. More details on the Bayes’ theorem
can be found in the Supporting Information. Apart from the physical
significance of the model and the individual model parameters, the
regression results were also evaluated on a statistical basis. The
F test for the global significance of the regression was performed.
The statistical significance of the parameter estimates was assessed
with the Student t-test. The binary correlation coefficients between
the estimated parameters were calculated [55–57].
N
N
H stretching band, which can be found at 3314 cm−1, while the
H stretching of phenylaminopropane (Fig. 2(5)) can be found at
3417 cm−1. Fig. 2(6) exhibits no N H stretching due to the absence
of a hydrogen on the nitrogen atom of diethylaminopropane. The
successful removal of the silanol groups after threatment with
1,1,1,3,3,3-hexamethyldisilazane (HMDS) is demonstrated by the
complete disappearence of the OH stretching band at 3745 cm−1
.
The active site concentrations, as determined from CHNS ele-
mental analysis, are reported in Table 2. The concentrations of
amine groups on cooperative acid–base (AB) as well as on unpro-
moted base (B) samples are calculated from the obtained nitrogen
weight percentage and are found to be identical. The free silanol
concentration on the cooperative acid–base samples is determined
from the increase in carbon after HMDS treatment, as one silanol
3. Experimental results
3.1. Catalyst characterization
A type IV nitrogen adsorption–desorption isotherm is obtained
for the calcined Silicagel 60 material. It has a specific BET surface
Please cite this article in press as: J. Lauwaert, et al., Effects of amine structure and base strength on acid–base cooperative aldol
condensation, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.08.007

G Model
CATTOD-9214; No. of Pages11
ARTICLE IN PRESS
J. Lauwaert et al. / Catalysis Today xxx (2014) xxx–xxx
5
Fig. 2. Typical DRIFT spectra of (1) pretreated silica; and amine functionalized catalysts using (2) APTES 6; (3) MAPTMS 3; (4) CAPTMS 2; (5) PAPTMS 1; (6) DEAPTMS 3; (a)
cooperative acid–base catalysts (type AB) and (b) unpromoted base catalysts (type B).
is capped by one trimethylsilyl moiety and, hence, is equivalent
with a mass increase corresponding to three carbon atoms. The
experimental error was found to be independent of concentration
and active site type and equal to 0.02 wt%. By applying the rules
of error propagation, it is found that the experimental error in the
molar silanol-to-amine ratio decreases with decreasing ratios.
In the presence of promoting silanol groups, the nucleophilic reac-
tions will be facilitated due to the formation of hydrogen-bridge
interactions between carbonyl moieties and the silanol groups, rep-
resented by the species indicated by (1a), (2a), (3a) and (4a).
In the following sections the effects of the presence of silanol
groups, the arrangement of the sites on the silica surface, the nature
of the amine active sites and the reaction conditions on the catalytic
performance is investigated.
3.2. Catalytic performance evaluation and interpretation
In our previous work [58], a reaction mechanism has been
proposed for the primary-amine-catalysed aldol condensation
between 4-nitrobenzaldehyde and acetone. Scheme 2 represents
a further elaborated version of this reaction mechanism which will
allow assessing the amine structure and base strength effects inves-
tigated as part of the present work. It shows the reaction steps
occurring in absence as well as in presence of promoting silanol
groups. When no promoting silanol groups are available, the first
step of the reaction mechanism is the formation of an enamine (1)
through a nucleophilic addition of acetone to the amine active site.
This enamine is in equilibrium with an imine (3) which inhibits the
further reaction. For the actual aldol condensation to take place,
the enamine reacts with 4-nitrobenzaldehyde yielding an iminium
ion (2). The primary reaction product (Aldol) is formed by a water-
assisted desorption of this iminium ion. Finally, the aldol product
releases water and forms the secondary reaction product (Ketone).
3.2.1. Operating conditions
The effect of the operating conditions on the catalyst perfor-
mance is investigated. Notably, the same trends were observed
for all types of active sites. The temperature effect is investi-
gated in a temperature range from 288 to 328 K. In accordance
with the Arrhenius law, a higher temperature results in a higher
turnover frequency. The selectivity towards the secondary ketone
product increases with increasing temperature. Also the initial
reactant concentrations are varied. The initial molar ratio of ace-
tone to 4-nitrobenzaldehyde is varied in the range of 15:1–500:1.
At low 4-nitrobenzaldehyde concentrations the turnover frequency
increases with increasing 4-nitrobenzaldehyde concentration,
however, above an initial 4-nitrobenzaldehyde to amine ratio of 50
the turnover frequency remains constant. In literature, a decreasing
reaction rate has been observed at high 4-nitrobenzaldehyde con-
centrations when using a primary amine as active site [22]. This
Please cite this article in press as: J. Lauwaert, et al., Effects of amine structure and base strength on acid–base cooperative aldol
condensation, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.08.007

G Model
CATTOD-9214; No. of Pages11
ARTICLE IN PRESS
6
J. Lauwaert et al. / Catalysis Today xxx (2014) xxx–xxx
Scheme 2. Proposed reaction mechanism for the aldol condensation of 4-nitrobenzaldehyde with acetone.
result suggested that an inhibiting imine could be formed between
a primary amine and 4-nitrobenzaldehyde. In the present work an
acetone excess has been used. As a result the direct interaction
between 4-nitrobenzaldehyde and the active sites is suppressed.
Correspondingly, also the formation of the inhibiting imine from
4-nitrobenzaldehyde can be considered limited [22,23]. The ace-
tone excess used also results in a rather limited effect of the initial
acetone concentration on the turnover frequency. Moreover, a vari-
ation in the initial concentrations of the reactions does not impact
on the selectivities towards the reaction products for a given con-
version.
TOF, while the highest turnover frequencies were obtained with
n-methylaminopropyltrimethoxysilane as precursor.
At high conversions a selectivity towards the secondary ketone
product of about 16% is obtained with the aminopropane active site
while a selectivity of about 10% is obtained with the methylamino-
propane active site. Hence, the nature of the active site seems to
influence the obtained product distribution.
3.2.2.1. The promoting effect of the silanol groups and the arrange-
ment of active sites. The data reported in Table 3 clearly show how
the activity exhibited by amine groups is enhanced by the presence
of silanols, which confirms the cooperativity between the acid and
base sites [27,28,58]. The product distribution is not significantly
affected by the presence of silanol groups, however.
3.2.2. Catalyst properties
In order to investigate the effect of the catalyst properties on the
aldol condensation, all catalysts were subject to identical operating
conditions. Pronounced differences in turnover frequency (TOF) are
observed as a function of the base that was grafted on the silica,
see Table 3. While primary amines exhibit a reasonable TOF, that
observed with tertiary amines is extremely low, i.e., in the noise
level. This is, however, consistent with literature reported results
[22,28]. The TOF obtained with secondary amines depended on the
substituent. Cyclohexyl and phenyl substituents both lead to a low
Recently we showed that the cooperativity does not only
depend on the ratio of both types of active sites but also on the
arrangement of these active sites on the catalyst surface [58].
Thermodynamic calculations for a n-propylamine–toluene mix-
ture revealed positive deviations from ideal liquid behaviour,
i.e., propylamine–propylamine interactions preferentially occur as
compared to propylamine–toluene interactions. Hydrogen bonding
between the primary amines is deemed to be responsible for this
and results in primary amines which are positioned in a clustered
manner on the silica surface. This is reflected in the percentage of
amines which is promoted by silanols as shown in Fig. 3. The same
figure also shows percentage of amines promoted by silanols when
using n-methylaminopropyltrimethoxysilane as precursor. It is evi-
dent from Fig. 3 that this precursor leading to a secondary amine
is grafted in a random manner on the silica surface. Thermody-
namic calculations for a n-methylaminopropane–toulene mixture
show no significant deviations from ideality. This indicates that
methylaminopropane–methylaminopropane interactions are sim-
ilar to methylaminopropane–toluene interactions and that the
precursor has no tendency to cluster in the synthesis mixture nor
on the silica surface. This also results in a higher percentage of
promoted secondary amines in the silanol-to-amine ratio range of
0–1.7 in comparison to the primary amines. Due to the low activi-
ties obtained using the other types of active sites, especially in the
Table 3
Turnover frequencies obtained with both the unpromoted base as the corresponding
cooperative acid–base catalysts.
a
b
Amine active site
TOFunprom (s−1
)
TOFprom (s−1
)
Aminopropane
2.0 × 10−4
1.0 × 10−3
9.6 × 10−6
–
7.8 × 10−4
3.3 × 10−3
3.1 × 10−5
7.0 × 10−6
3.9 × 10−5
Methylaminopropane
Cyclohexylaminopropane
Phenylaminopropane
Diethylaminopropane
1.2 × 10−5
a
Turnover frequencies calculated by averaging the TOFs obtained using the
unpromoted base catalysts (samples B).
b
Turnover frequencies calculated by extrapolating the TOFs obtained using the
cooperative acid–base catalysts (samples AB) towards a molar silanol-to-amine ratio
of infinity.
Please cite this article in press as: J. Lauwaert, et al., Effects of amine structure and base strength on acid–base cooperative aldol
condensation, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.08.007

G Model
CATTOD-9214; No. of Pages11
ARTICLE IN PRESS
J. Lauwaert et al. / Catalysis Today xxx (2014) xxx–xxx
Table 4
7
ꢁH◦, ꢁS◦ and corresponding pKa,298K values for deprotonation [59].
Amine active site
ꢁH◦
ꢁS◦
pKa,298K of the
(kJ/mol)
(J/mol K)
ammonium ion
Aminopropane
57.91
52.59
57.49
33.39
43.43
−8.37
−30.12
−17.15
3.56
10.58
10.79
10.97
5.66
Methylaminopropane
Cyclohexylaminopropane
Phenylaminopropane
Diethylaminopropane
−60.67
10.78
reaction with the formation of the aldol product. However, except
for phenylaminopropane, the differences in basicity of the amines
are limited. The low basicity for molecules such as phenylamino-
propane is a result from the stabilization of the nitrogen lone pair of
electrons by conjugation with the aromatic ring ␲ system. This low
base strength limits the nucleophilic reactions and explains the low
TOF obtained with this active site. However, the limited differences
in basicity of the other amines do not explain the experimentally
observed TOFs and, hence, other factors, such as the inherent pos-
sibility to form the reactive enamine intermediate or an inhibiting
imine on the catalyst surface, may exert a dominant effect.
Fig. 3. Percentage of amines promoted as a function of the silanol-to-amine ratio;
ꢀcatalysts functionalized using APTES; ꢁ catalysts functionalized using MAPTMS;
dotted line calculated percentage by means of randomly generated surface; dashed
line calculated percentage accounting for clustering in the synthesis mixture [58].
low silanol-to-amine ratio range, no conclusive experimental evi-
dence could be obtained about the arrangement of these types of
amines on the silica surfaces.
3.2.2.3. Amine structure effects. As shown in Scheme 2 the first step
in the aldol condensation reaction mechanism is the nucleophilic
addition of acetone to the amine group with concomitant forma-
tion of an enamine. This reaction happens in two steps. First a
carbinolamine is formed which, subsequently, dehydrates with for-
mation of the enamine, see Scheme 3. In the case of a primary amine
the carbinolamine can also dehydrate with formation of an imine,
which is an inhibiting species on the catalyst surface. The acetone
excess used, makes that the imine corresponding with acetone is
the only one which is formed whereas the formation of the one
3.2.2.2. Base strength. Table 4 shows the base strength of the amine
active sites, expressed as pKa of the corresponding ammonium ion
at 298 K, that have been used in the present work. These values were
obtained from the enthalpy and entropy values calculated using the
group contribution method described by Christensen et al. [59]. It
can be expected that the base strength influences all nucleophilic
reactions occurring on the amine site itself, i.e., the nucleophilic
addition of acetone with the formation of the enamine interme-
diate (1), the formation of the ketone product and the reverse
Scheme 3. Nucleophilic addition of acetone to (a) a primary amine and (b) a secondary amine.
Please cite this article in press as: J. Lauwaert, et al., Effects of amine structure and base strength on acid–base cooperative aldol
condensation, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.08.007

G Model
CATTOD-9214; No. of Pages11
ARTICLE IN PRESS
8
J. Lauwaert et al. / Catalysis Today xxx (2014) xxx–xxx
Table 5
UNIQUAC binary interaction parameters estimated based upon COSMO-RS calculations and corresponding 95% confidence intervals.
j
i
Acetone
n-Hexane
ꢁuij (J/mol)
ꢁuji (J/mol)
ꢁuij (J/mol)
ꢁuji (J/mol)
Acetone
4-Nitrobenzaldehyde
Aldol
Ketone
Water
0.0
0.0
−71 42
220 78
660 101
378 102
756 130
1710 57
1510 91
1690 113
1230 109
1450 123
−979 152
3470 158
1970 181
−1080 178
560 213
−2500 41
−1590 85
899 282
corresponding with 4-nitrobenzaldehyde is suppressed [22,23]. In
case of a secondary amine the formation of any imine is prevented
due to the lack of a second hydrogen on the nitrogen. For a simi-
lar carbinolamine concentration, this results in a higher enamine
concentration when the active site is a secondary amine compared
to the case that the active site is a primary amine. Hence, also a
higher reaction rate for the following reaction steps is obtained in
the case of secondary amine. In case of a tertiary amine the forma-
tion of the carbinolamine intermediate is no longer possible due to
the absence of hydrogen atoms on the nitrogen atom, explaining
the low activity of the tertiary amine.
described by a composite reaction step. Moreover, also the interme-
diates in the dehydration of the aldol product to form the ketone
product and the backward reaction are not explicitly considered.
This is done in order to reduce the amount of parameters to be
estimated during the regression. The reactions resulting in the for-
mation of hydrogen-bridge interactions between carbonyl moieties
and the silanol groups (1a), (2a), (3a) and (4a) and the reactions
between the enamine (1) and the inhibiting imine (3) are consid-
ered to be quasi-equilibrated while all other reactions are taken as
irreversible forward reactions.
If the amine is promoted by the presence of a silanol, the net rate
of the enamine (1) formation is given by Eq. (2). In this equation aace
and abenz, represent, respectively, the activity of acetone and the
activity of 4-nitrobenzaldehyde. In case of an unpromoted amine,
both the first and the fourth term of Eq. (2) disappear. The reaction
rate coefficients are described by the Arrhenius law, which is repa-
rameterized in order to avoid strong binary correlation between
the pre-exponential factor and the activation energy [60].
The cyclohexylaminopropane site has a low activity, although it
has the highest base strength and it is unable to from the inhibiting
imine. This can be explained by steric hindrance induced by the
bulky cyclohexyl group close to the amine.
4. Kinetic model based analysis of the amine catalysed
aldol condensation
In this work, the primary amine is considered to be the reference
active site. Changing this active site to another type of amine has an
effect on all the reaction steps occurring on the amine itself, i.e., the
enamine (1) formation, the desorption of the intermediate (2), the
imine (3) formation, the product ketone formation and the reverse
reaction with the product aldol formation. As already described in
Section 3.2.2, the imine can only be formed when the active site is
a primary amine. In all other cases the equilibrium coefficient Kim
equals zero. The effect on the other reaction steps cited above is
described by a change in activation entropy ꢁSamine and activation
energy ꢁEamine between the reference primary amine and the con-
sidered active site. For reaction rate coefficient k1 this is shown in
Eq. (3). The changes in reaction entropy and reaction enthalpy are
assumed to be identical for all affected reaction rate coefficients, i.e.,
k1, k3, k4 an k–4. The presence of silanol groups can have a twofold
effect on the reaction enthalpy of the four adsorption steps, namely
a first one due to the prior adsorption of the reactant on a silanol
group and a second one which is related to the higher susceptibil-
ity of species (1a) and (2a) to nucleophilic reactions compared to
bulk acetone and 4-nitrobenzaldehyde (Eq. (4)). As it is impossible
to distinguish between these two effects based on the experimen-
tation performed as part of the present work, both effects have
4.1. Interaction parameter estimation and activity coefficient
calculation
The activity coefficients which were determined using COSMO-
RS are used to estimate the binary interaction parameters of the
UNIQUAC model. The obtained F value for the global significance
of the regression amounts to 7476, which exceeds the tabulated
value by 4 orders of magnitude. Hence, the model behaviour can
be considered to be acceptable. The binary interaction parameters
and corresponding 95% confidence intervals are shown in Table 5. A
figure showing the effects of the liquid composition and tempera-
ture on the activity coefficients has been included in the Supporting
Information.
The activity coefficient of water is the most sensitive to both
the composition and the temperature of the liquid and deviates
most significantly from the ideal value of 1. If acetone is the most
abundant component in the liquid, the activity coefficients of 4-
nitrobenzaldehyde and the ketone product are close to 1, while
the activity coefficient of the aldol product is only about 0.5. If
n-hexane is the most abundant component in the liquid, the activ-
ity coefficients of 4-nitrobenzaldehyde, the aldol product and the
ketone product increase to values around, respectively, 2.5, 4.0
and 3.0. Moreover, in this case, the effect of the mole percent-
age of the component in the liquid on the activity coefficient is
more pronounced. In the range of investigated experimental con-
ditions, the activity coefficient of acetone is not much affected by
the presence of other components except for n-hexane. An increas-
ing amount of n-hexane leads to an increasing activity coefficient
of acetone.
ꢀ
been combined into a single parameter ꢁEOH . The adsorption on
a silanol group induces also a change in the reaction entropy, i.e.,
ꢀ
ꢁSOH . Also in this case, the shift in reaction entropy and reaction
enthalpy is assumed to be identical for all adsorption steps:
ꢀ
R1,prom = k1 Kads,1C0,promaace + k1C0,promaace − k2C1,prom
ꢀ
− k2
K
ads,2C1,promabenz
(2)
(3)
k1 = A1,ref e−ꢁS
/Re((−(E
+ꢁE
))/R)((1/T)−(1/T ))
ref
1
amine
amine
4.2. Kinetic model construction and descriptor identification
−ꢁH
)/R)((1/T)−(1/T ))
k1 Kads,1 = k1e(−ꢁS
/R)e((ꢁE
OH
ads,1
ads,1 ref
ꢀ
The reaction mechanism shown in Scheme 2 is used for kinetic
model construction. Note that, the carbinolamine intermediate is
not explicitly accounted for because the enamine (1) formation is
(ꢁE
/R)((1/T)−(1/T ))
= k1e−ꢁS /Re
(4)
ꢀ
ꢀ
OH
ref
OH
Please cite this article in press as: J. Lauwaert, et al., Effects of amine structure and base strength on acid–base cooperative aldol
condensation, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.08.007

G Model
CATTOD-9214; No. of Pages11
ARTICLE IN PRESS
J. Lauwaert et al. / Catalysis Today xxx (2014) xxx–xxx
9
Fig. 4. (a) ꢁSamine as a function of ꢁS◦ for deprotonation and (b) ꢁEamine as a function of ꢁH◦ for deprotonation; the dashed lines are intended as visual aids only.
4.3. Determination of the descriptors and discussion of the
reaction mechanism
present these activation energies decrease by about 9 kJ/mol due to
the formation of a hydrogen bond between the silanol and the reac-
tant. This leads to an apparent activation energy for the formation
of the enamine intermediate which is very low. The formation of
a hydrogen bond between a silanol and a carbonyl moiety induces
an entropy decrease of about 8 J/(mol K).
The model performance is visualized by the parity diagrams
in Fig. 5. Independently of type of catalyst, i.e., unpromoted or
promoted and primary or secondary amine, the simulated concen-
trations of the reactants and both reaction products agree well with
the observed concentrations.
Using the estimated parameters the model shows that at the
operating conditions used to investigate the effect of the catalytic
properties, similar amounts of free active sites (0) and inhibiting
imines (3) occur on the surface of a fully promoted primary amine
catalyst, i.e., respectively 43% and 45%. The remainder of the surface
is mainly occupied by the enamine intermediate (1). The iminium
ion (2) occupies less than 1% of the surface. It is also shown for a fully
promoted primary amine catalyst used at these conditions that 74%
of the reactants are transformed via the promoted pathway, i.e., via
intermediates (1a) and (2a), while the other 26% of the reactants
are transformed without the aid of a silanol group.
The effect of the nature of the amine active site is assessed
by a change in activation entropy and activation energy between
the reference primary amine and the considered active site. In
Fig. 4 these changes are plotted as a function of the reaction
entropies and enthalpies for deprotonation of the corresponding
active sites, i.e., amine base strength. The decrease of the activation
entropy induced by changing the active site exhibits an apparent
proportionality to the decrease in reaction entropy for deproto-
nation while for the activation energies no such proportionality is
observed. When the sterically hindered secondary amine (CAPTMS)
is used as active site the activation energies are about 17 kJ/mol
higher than expected from the results obtained with the other two
amines. This high activation energy can be related to the steric hin-
drance induced by the bulky cyclohexyl group close to the amine.
For the accessible secondary amine (MAPTMS) type of cata-
lysts the model shows that at the operating conditions used to
Due to the very low activity of phenylaminopropane resulting
from a low base strength and the inability of diethylaminopropane
to form the carbinolamine intermediate, a kinetic model assess-
ment of the aldol condensation kinetics is only relevant for the
catalysts obtained by grafting APTES, MAPTMS and CAPTMS, i.e.,
a primary amine, an accessible secondary amine and a sterically
hindered secondary amine.
The equilibrium coefficient Kim and the reaction enthalpy of the
reaction forming the secondary ketone product, which is equal to
difference between the activation energies E4 and E–4, are deter-
mined using the Joback method [53], while the other parameters
are estimated by means of regression. The obtained parameter val-
ues and the corresponding 95% confidence intervals are reported
in Tables 6 and 7. Notably, the rate coefficient of the water-
assisted desorption exceeds the other rate coefficients by far. As
a result of this the exact value of the rate coefficient does not
have a large influence on the obtained results and this rate coef-
ficient is non-significantly affected by the temperature, i.e., the
corresponding activation energy E3 is set equal to zero. The other
obtained activation energies are in the range of 11–51 kJ/mol which
is rather low for a chemical reaction, however, this is a result of the
composite character of these steps, see above. If a silanol group is
Table 6
Estimated values for the kinetic descriptors with their 95% confidence intervals
obtained after regression of the kinetic model (see Section 4.2) to the experimental
data (see Section 3.2).
Pre-exponential factor
Activation energy
k1
k2
k3
k4
k–4
5.04 × 10−5
5.81 × 10−2
0.23 × 10−5 l/(mol s)
11.7 1.4 kJ/mol
34.1 1.7 kJ/mol
0.0 kJ/molb
0.14 × 10−2 l/(mol s)
20.4 9.6 l/(mol s)
2.02 × 10−2
0.37 × 10−2 l/(mol s)
39.8 5.8 kJ/mol
4.60 1.12 l2/(mol2 s)
50.6 kJ/mola
a
Value determined using the Joback method.
Non-significantly estimated parameter.
b
Table 7
Estimated catalyst descriptors with their 95% confidence intervals obtained after regression of the kinetic model (see Section 4.2) to the experimental data (see Section 3.2).
Entropic difference
Reaction enthalpy/activation energy difference
Promotional effect of silanols
ꢀ
ꢀ
ꢁSOH
−8.25 0.16 J/(mol K)
ꢁEOH
8.7 1.8 kJ/mol
Nature of the amine sites
ꢁSim
13.5 J/(mol K)a
ꢁHim
43.8 kJ/mola
ꢁSMAPTMS
ꢁSCAPTMS
−46.4 2.6 J/(mol K)
ꢁEMAPTMS
ꢁECAPTMS
10.4 1.0 kJ/mol
17.6 3.2 kJ/mol
−23.7 5.0 J/(mol K)
a
Value determined using the Joback method.
Please cite this article in press as: J. Lauwaert, et al., Effects of amine structure and base strength on acid–base cooperative aldol
condensation, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.08.007

G Model
CATTOD-9214; No. of Pages11
ARTICLE IN PRESS
10
J. Lauwaert et al. / Catalysis Today xxx (2014) xxx–xxx
Fig. 5. Parity diagrams for the responses of the kinetic model for aldol condensation on ♦ APTES catalysts; ꢂ MAPTMS catalysts and ꢁ CAPTMS catalysts: (1) acetone; (2)
4-nitrobenzaldehyde; (3) aldol and (4) ketone.
investigate the effect of the catalytic properties, the surface of a
fully promoted catalyst is mainly covered with the enamine inter-
mediate (1). Leaving only 1% of active sites free (0) which results in a
lower production of the unsaturated ketone product. The sterically
hindered secondary amine (CAPTMS) sites are mainly free (0). The
percentage of reactants which are transformed via the promoted
pathway remains the same as with the primary amine (APTES)
reference catalysts.
enamine intermediate without inhibiting imine formation can
exhibit the highest TOFs, provided that it is not subject to sterical
hindrance or unfavourable basicity effects. The latter is quantified
via differences in activation entropies and energies of all the reac-
tion steps occurring on the amine sites themselves, i.e., the enamine
formation, the water assisted desorption with the product aldol
formation, the product ketone formation and the reverse reaction
with product aldol formation. A low base strength, such as that of
phenylaminopropane results in a low TOF. A final factor which has
been considered in this work is steric hindrance. A cyclohexyl sub-
stituted on the amine results in an increase in activation energy
of about 17 kJ/mol. Hence, a secondary amine with a small sub-
stituent avoiding steric hindrance seems to be the optimal amine
type to catalyse the aldol condensation between acetone and 4-
nitrobenzaldehyde.
5. Conclusions
Distinct differences in turnover frequency (TOF) were obtained
when different types of amines were used to catalyse the aldol con-
densation between acetone and 4-nitrobenzaldehyde. The activity
of all amine active sites could be enhanced by incorporating silanol
groups in the catalyst. In contrast with primary amine active sites
which were found to be positioned in a clustered manner, sec-
ondary and tertiary amine active sites were randomly distributed
over the silica surface.
Acknowledgments
This work was supported by the Long Term Structural
Methusalem Funding by the Flemish Goverment. DE is a postdoc-
toral researcher of the FWO-Vlaanderen (Fund Scientific Research
– Flanders) grant number 3E10813W. The authors would like to
thank Tom Planckaert of the department of Inorganic and Physical
Chemistry for nitrogen sorption measurements and CHNS elemen-
tal analysis.
While primary amines exhibited
a reasonable TOF, that
observed with tertiary amines was extremely low. The TOF
obtained with secondary amines depended on the substituent.
Cyclohexyl and phenyl substituents both lead to a low TOF, while
the highest turnover frequencies among all investigated catalyst
were obtained secondary amine active sites with a methyl sub-
stituent. Activity differences are explained by an interplay of the
following three factors, i.e., the possibility of forming a reac-
tive enamine intermediate and inhibiting imine species with the
amine active site, the base strength of the amine active site and
steric hindrance effects. The possibility to form a reactive inter-
mediate and inhibiting species was the most important factor
determining the observed TOFs. While the crucial enamine reactive
intermediate cannot be formed on a tertiary amine, the inhibit-
ing imine can only be formed on primary amine active sites. As a
result, a secondary amine active site which can yield the reactive
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.cattod.2014.08.007.
References
[1] C. Lucarelli, A. Vaccari, Green Chem. 13 (2011) 1941.
Please cite this article in press as: J. Lauwaert, et al., Effects of amine structure and base strength on acid–base cooperative aldol
condensation, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.08.007

G Model
CATTOD-9214; No. of Pages11
ARTICLE IN PRESS
J. Lauwaert et al. / Catalysis Today xxx (2014) xxx–xxx
11
[2] K. Tanabe, W.F. Holderich, Appl. Catal. A-Gen. 181 (1999) 399.
[3] S. Katare, J.M. Caruthers, W.N. Delgass, V. Venkatasubramanian, Ind. Eng. Chem.
Res. 43 (2004) 3484.
[4] K. Metaxas, J.W. Thybaut, G. Morra, D. Farrusseng, C. Mirodatos, G.B. Marin, Top.
Catal. 53 (2010) 64.
[33] S. Huh, H.T. Chen, J.W. Wiench, M. Pruski, V.S.Y. Lin, Angew. Chem. Int. Ed. 44
(2005) 1826.
[34] N. Solin, L. Han, S. Che, O. Terasaki, Catal. Commun. 10 (2009) 1386.
[35] B. Voit, Angew. Chem. Int. Ed. 45 (2006) 4238.
[36] F.P. Shang, J.R. Sun, S.J. Wu, Y. Yang, Q.B. Kan, J.Q. Guan, Micropor. Mesopor.
Mater. 134 (2010) 44.
[37] S. Shylesh, A. Wagner, A. Seifert, S. Ernst, W.R. Thiel, Chem. Eur. J. 15 (2009)
7052.
[38] J.D. Bass, A. Solovyov, A.J. Pascall, A. Katz, J. Am. Chem. Soc. 128 (2006) 3737.
[39] P. Van Der Voort, E.F. Vansant, J. Liq. Chromatogr. Relat. Technol. 19 (1996)
2723.
[40] L.T. Zhuravlev, Colloids Surf. A-Physicochem. Eng. Asp. 74 (1993) 71.
[41] L.T. Zhuravlev, Colloids Surf. A-Physicochem. Eng. Asp. 173 (2000) 1.
[42] P. Van Der Voort, S. Vercauteren, K. Peeters, E.F. Vansant, J. Colloid Interface Sci.
157 (1993) 518.
[5] I.R. Choudhury, K. Hayasaka, J.W. Thybaut, C.S.L. Narasimhan, J.F. Denayer, J.A.
Martens, G.B. Marin, J. Catal. 290 (2012) 165.
[6] K. Weissermel, H.-J. Arpe, Industrial Organic Chemistry, Wiley, New York, 1997.
[7] Ullmann’s Encyclopedia of Industrial Chemistry, 2008.
[8] C. Chapuis, D. Jacoby, Appl. Catal. A-Gen. 221 (2001) 93.
[9] S. Bhaduri, D. Mukesh, Homogeneous Catalysis: Mechanisms and Industrial
Applications, John Wiley & Sons, Inc., New York, 2000.
[10] J.N. Chheda, G.W. Huber, J.A. Dumesic, Angew. Chem. Int. Ed. 46 (2007) 7164.
[11] A. Wolfson, G. Litvak, C. Dlugy, Y. Shotland, D. Tavor, Ind. Crop. Prod. 30 (2009)
78.
[12] J.C. Serrano-Ruiz, J.A. Dumesic, Energy Environ. Sci. 4 (2011) 83.
[13] R.E. O’Neill, L. Vanoye, C. De Bellefon, F. Aiouache, Appl. Catal. B-Environ. 144
(2014) 46.
[43] E.F. Vansant, P. Van Der Voort, Characterization and Chemical Modification of
the Silica Surface, 2 Ed., Studies in Surface Science and Catalysis, 1997.
[44] M.E. Davis, R.J. Davis, Fundamentals of Chemical Reaction Engineering,
McGraw-Hill, New York, 2003.
[45] B.E. Poling, J.M. Prausnitz, J.P. O’Connell, The Properties of Gases and Liquids,
McGraw-Hill, New York, 2001.
[46] D.S. Abrams, J.M. Prausnitz, AICHE J. 21 (1975) 116.
[47] A. Fredenslund, R.L. Jones, J.M. Prausnitz, AICHE J. 21 (1975) 1086.
[48] H.K. Hansen, P. Rasmussen, A. Fredenslund, M. Schiller, J. Gmehling, Ind. Eng.
Chem. Res. 30 (1991) 2352.
[14] M.J. Climent, A. Corma, S. Iborra, Green Chem. 16 (2014) 516.
[15] N.C. Nelson, U. Chaudhary, K. Kandel, I.I. Slowing, Top. Catal. 57 (2014) 1000.
[16] W. Kagunya, W. Jones, Appl. Clay Sci. 10 (1995) 95.
[17] M. Sahoo, S. Singha, K.M. Parida, New J. Chem. 35 (2011) 2503.
[18] K.K. Rao, M. Gravelle, J.S. Valente, F. Figueras, J. Catal. 173 (1998) 115.
[19] E. Suzuki, Y. Ono, Bull. Chem. Soc. Jpn. 61 (1988) 1008.
[20] J. Roelofs, A.J. van Dillen, K.P. de Jong, Catal. Today 60 (2000) 297.
[21] C.L. Xu, Y. Gao, X.H. Liu, R.R. Xin, Z. Wang, RSC Adv. 3 (2013) 793.
[22] K. Kandel, S.M. Althaus, C. Peeraphatdit, T. Kobayashi, B.G. Trewyn, M. Pruski,
I.I. Slowing, J. Catal. 291 (2012) 63.
[49] A. Klamt, J. Phys. Chem. 99 (1995) 2224.
[50] A. Klamt, F. Eckert, Fluid Phase Equilib. 172 (2000) 43.
[51] F. Eckert, A. Klamt, AICHE J. 48 (2002) 369.
[23] K. Kandel, S.M. Althaus, C. Peeraphatdit, T. Kobayashi, B.G. Trewyn, M. Pruski,
I.I. Slowing, ACS Catal. 3 (2013) 265.
[24] N.A. Brunelli, S.A. Didas, K. Venkatasubbaiah, C.W. Jones, J. Am. Chem. Soc. 134
(2012) 13950.
[52] A. Klamt, F. Eckert, Fluid Phase Equilib. 217 (2004) 53.
[53] K.G. Joback, R.C. Reid, Chem. Eng. Commun. 57 (1987) 233.
[54] W.E. Stewart, M. Caracotsios, Computer-Aided Modeling of Reactive Systems,
Wiley-Interscience, Hoboken, NJ, 2008.
[25] N.A. Brunelli, K. Venkatasubbaiah, C.W. Jones, Chem. Mater. 24 (2012) 2433.
[26] N.A. Brunelli, C.W. Jones, J. Catal. 308 (2013) 60.
[55] G.F. Froment, L.H. Hosten, in: J.R. Anderson, M. Boudart (Eds.), Catalysis Science
and Technology, Springer Verlag, Berlin, 1981.
[27] Y. Kubota, K. Goto, S. Miyata, Y. Goto, Y. Fukushima, Y. Sugi, Chem. Lett. 32
(2003) 234.
[28] Y. Kubota, H. Yamaguchi, T. Yamada, S. Inagaki, Y. Sugi, T. Tatsumi, Top. Catal.
53 (2010) 492.
[29] S.F. Cheng, X.G. Wang, S.Y. Chen, Top. Catal. 52 (2009) 681.
[30] Y.Q. Shao, J.Q. Guan, S.J. Wu, H. Liu, B. Liu, Q.B. Kan, Micropor. Mesopor. Mater.
128 (2010) 120.
[31] R.K. Zeidan, S.J. Hwang, M.E. Davis, Angew. Chem. Int. Ed. 45 (2006) 6332.
[32] R.K. Zeidan, M.E. Davis, J. Catal. 247 (2007) 379.
[56] L.H. Hosten, G.F. Froment, in: L.K. Doraiswamy (Ed.), Recent Advances in the
Engineering Analysis of Chemically Reaction Systems, Wiley Eastern, New
Delhi, 1984, p. 68.
[57] Y. Bard, Nonlinear Parameter Estimation, Academic Press, New York, 1974.
[58] J. Lauwaert, E. De Canck, D. Esquivel, J.W. Thybaut, P. Van Der Voort, G.B. Marin,
ChemCatChem 6 (2014) 255.
[59] J.J. Christensen, R.M. Izatt, D.P. Wrathall, L.D. Hansen, J. Chem. Soc. A-Inorg.
Phys. Theoret. (1969) 1212.
[60] M. Schwaab, J.C. Pinto, Chem. Eng. Sci. 62 (2007) 2750.
Please cite this article in press as: J. Lauwaert, et al., Effects of amine structure and base strength on acid–base cooperative aldol
condensation, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.08.007