Amino-functionalized Zr(IV) metal-organic framework as bifunctional acid-base catalyst for Knoevenagel condensation

Article abstract of DOI:10.1016/j.molcata.2014.04.002

The amino-functionalized metal-organic framework of Zr(IV) with 2-aminoterephthalate, UiO-66-NH₂, was studied as a solid catalyst for Knoevenagel condensation. The material can efficiently catalyze the condensation reaction of benzaldehyde with

Full text of DOI:10.1016/j.molcata.2014.04.002

Accepted Manuscript
Title: Amino-functionalized Zr(IV) metal-organic framework
as bifunctional acid-base catalyst for Knoevenagel
condensation
Author: Yang Yang Hong-Fei Yao Fu-Gui Xi En-Qing Gao
PII:
S1381-1169(14)00140-X
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.molcata.2014.04.002
MOLCAA 9064
To appear in:
Journal of Molecular Catalysis A: Chemical
Received date:
Revised date:
Accepted date:
6-12-2013
1-4-2014
3-4-2014
Please cite this article as: Y. Yang, H.-F. Yao, F.-G. Xi, E.-Q. Gao, Amino-
functionalized Zr(IV) metal-organic framework as bifunctional acid-base catalyst
for Knoevenagel condensation, Journal of Molecular Catalysis A: Chemical (2014),
http://dx.doi.org/10.1016/j.molcata.2014.04.002
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

*Graphical Abstract (for review)
Page 1 of 37

*Highlights (for review)
Highlights
 UiO-66-NH2 catalyzes the reaction of aldehydes with cyanoacetate or
malononitrile.
 The catalyst is heterogeneous, recyclable and shows size effects on substrates
 It is proposed that the activity may arise from a dual acid–base character.
Page 2 of 37

Amino-functionalized Zr(IV) metal-organic framework as bifunctional
acid-base catalyst for Knoevenagel condensation
Yang Yang, Hong-Fei Yao, Fu-Gui Xi and En-Qing Gao*
Shanghai Key Laboratory of Green Chemistry and Chemical Processes, Department of
Chemistry, East China Normal University
Shanghai 200062, China
main text with abstract, keywords, references and embedded Tables and figures: pages 1-22
captions for Schemes and Figures: page 23
2 tables: pages 24, 25
2 schemes: pages 26, 27
6 figures: pages 28-34
Corresponding author:
En-Qing Gao
Shanghai Key Laboratory of Green Chemistry and Chemical Processes, Department of
Chemistry, East China Normal University
Shanghai 200062, China
Fax: +86-21-62233404
E-mail: eqgao@chem.ecnu.edu.cn
0
Page 3 of 37

Amino-functionalized Zr(IV) metal-organic framework as bifunctional
acid-base catalyst for Knoevenagel condensation
Yang Yang, Hong-Fei Yao, Fu-Gui Xi and En-Qing Gao*
Shanghai Key Laboratory of Green Chemistry and Chemical Processes, Department of
Chemistry, East China Normal University, Shanghai 200062, China. E-mail:
eqgao@chem.ecnu.edu.cn
Abstract:
The amino-functionalized metal-organic framework of Zr(IV) with 2-aminoterephthalate,
UiO-66-NH2, was studied as a solid catalyst for Knoevenagel condensation. The material can
efficiently catalyze the condensation reaction of benzaldehyde with ethyl cyanoacetate or
malononitrile in highly polar solvents such as DMF, DMSO and ethanol. The catalytic system
has also been tested for various aromatic aldehydes, the conversion easily reaching more than
90% under mild conditions. It was demonstrated that the catalytic process is heterogeneous
and shows size effects, characteristic of a porous catalyst. The catalyst can be recycled
without losing its framework integrity and catalytic activity. The catalytic activity has been
compared with dimethyl 2-aminoterephthalate and the isostructural amino-free MOF
(UiO-66). The superior performance of UiO-66-NH2 has been attributed to the site-isolated
acid-base bifunctional character. It has been proposed that the Zr site in close proximity to the
amino group activates aldehydes to promote the formation of aldimine intermediates from the
aldehydes and the amino group.
Keywords: Metal-organic frameworks; Bifunctional acid-base catalysts; Knoevenagel
condensation; Heterogeneous catalysis
1
Page 4 of 37

1.Introduction
In recent years, metal–organic frameworks (MOFs) have attracted considerable attention
due to their special characteristics such as hybrid compositions, diverse networks, tunable
porosity and tailorable surfaces [1, 2]. Owing to these characteristics, MOFs are promising
materials for various technological applications such as gas storage/capture [3-5], separation
[6, 7], and heterogeneous catalysis [8-11]. Heterogeneous catalysis is superior to
homogeneous catalysis for easier separation, reusability, minimized waste and cleaner
products. Heterogeneous catalysts also offer the possibility of combining isolated acidic and
basic sites for cooperative or tandem catalysis [12-15]. Furthermore, porous solid catalysts
may provide confined space to influence reactivity and selectivity. The application of MOFs
as porous heterogeneous catalysts alternative or complementary to microporous zeolites is
especially interesting, since the pore size and chemical functionality of MOFs can be
modulated within a wider range [2, 11, 16]. In principle, the active sites of MOF catalysts can
be metal centers with unsaturated (or labile) coordination environments, catalytic functional
groups attached to any components of the frameworks, or other catalytic species (molecules,
metals, etc.) encapsulated in the pores, and the sites may be inherent in the frameworks or
generated by post-synthetic methods [11, 17-22].
The three-dimensional (3D) Zr(IV) MOFs of formula [Zr6O4(OH)4(L)6], such as UiO-66 [L
= 1,4-benzenedicarboxylate (BDC)] and UiO-66-NH2 [L = 2-amine-1,4-benzenedicarboxylate
(BDC-NH2)], are based on the octahedral [Zr6(µ3-O)4(µ3-OH)4(µ2-COO)12] cluster and
featured by alternating octahedral and tetrahedral cages sharing triangular windows [23] (Fig.
1). The MOFs contain potential Lewis-acid sites (Zr(IV) centers) upon activation and/or base
sites (NH2 groups), and remarkably, they show exceptionally high thermal and chemical
stability. However, for only a few reactions have these MOFs been studied as heterogeneous
catalysts, including the cross-aldol condensation between benzaldehyde and heptanal [24], the
cyclization of citronellal to isopulegol [25, 26], the cycloaddition of CO2 to styrene oxide [27],
the acetalization of benzaldehyde with methanol [28], and photocatalytic reactions [29-34].
Electronic effects of linker substitution on Lewis acid catalysis have been studied [28]. It was
found that the Lewis acidity and the catalytic activity increase in the order of UiO-66-NH2 <
UiO-66 < UiO-66-NO2. It was also demonstrated that the introduction of amino groups into
2
Page 5 of 37

linkers leads to an increase in basicity [25, 28].
Fig. 1. Structure of UiO-66. (a) A [Zr6(µ3-O)4(µ3-OH)4(µ2-COO)12] cluster. (b) The porous
systems in which octahedral and tetrahedral cages share triangular windows.
Scheme 1. Knoevenagel condensation.
In this paper, we report the catalytic study of UiO-66-NH2 for Knoevenagel condensation.
The Knoevenagel condensation of a carbonyl group with the methylene group activated by
two electron withdrawing groups (Scheme 1) is an important C-C bond coupling reaction and
has been widely used in the synthesis of fine chemicals and pharmaceuticals. The
condensation can be catalyzed by bases or Lewis acids, either homogeneous (such as amines
[35, 36], ZnCl2 and Mg(ClO4)2 [37, 38]) or heterogeneous (such as hydrotalcite [39],
silica-supported basic catalysts [40, 41]). Considering the demand for environmentally
friendly heterogeneous catalysts and the intriguing features of MOFs, the exploration of MOF
catalysts for the Knoevenagel condensation has become a topic of increasing interest in recent
years [42-50]. The MOFs used are diverse in composition and structure, and most of them
contain the amino group as Lewis base sites, such as the BDC-NH2-base metal-carboxylate
frameworks IRMOF-3, MIL-101(Al)-NH2, MIL-53-NH2, and UMCM-1-NH2 [45, 46].
MIL-101(Cr) with post-synthetically introduced amino groups has also been studied for
Knoevenagel condensation [51]. In this work, we demonstrate that UiO-66-NH2 is an efficient,
size-selective, stable, and recyclable heterogeneous catalyst and that the superior catalytic
performance to UiO-66 is attributable to the bifunctional acid-base character.
3
Page 6 of 37

2. Experimental
2.1 Synthesis
All chemicals (A. R. grade, purity 98 wt% or higher) were obtained from commercial
sources and used without further purification. UiO-66-NH2 and UiO-66 synthesized by
following a solvothermal procedure reported in the literature [52, 53], using acetic acid as
modulator. ZrCl4 (0.7 mmol, 0.164 g) and H2BDC-NH2 (0.7 mmol, 0.127 g) were dissolved in
the mixture of N,N'-dimethylformamide (DMF, 8 ml), acetic acid (1.2 ml) and deionized
water (0.05 ml) at room temperature. After stirring for about 5 minutes, the mixture was
sealed in a 23 ml Teflon liner, heated in an oven at 120 ˚C for 48 hours, and then cooled to
room temperature. A yellow solid powder was obtained by filtration, washed three times with
DMF. The product was soaked and stirred in refluxing methanol for 24 h, filtered and dried at
70 ˚C in an oven. The formula of the material thus obtained was estimated to be
[Zr6O4(OH)4(BDC-NH2)6]14H2O according to elemental analysis. Found: C 29.2, H 3.8, N
4.4%; calculated: C 28.7, H 3.1, N 4.2%.
2.2 Characterization
The X-ray powder diffraction patterns of the samples were measured using a Rigaku
Ultima IV X-ray Diffractometer with Cu Kα radiation (λ = 1.54 Å) at a scanning rate of 10
˚C/min, with accelerating voltage and current of 35 kV and 25 mA, respectively. SEM study
was carried out with a S-4800 HITACHI scanning electron microscope. Elemental analyses
were determined on an Elementar Vario ELIII analyzer. FT-IR spectra were recorded in the
range 5004000 cm1 using KBr pellets on a Nicolet NEXUS 670 spectrophotometer. The
leached metal amount was detected by inductively coupled plasma atomic emission
spectroscopy (ICP-AES) using a IRIS Intrepid II XSP spectrometer. Gas chromatography (GC)
was conducted using a Linghua GC 9890E instrument equipped with an FID detector and an
SE-54 capillary column (30 m0.25 mm0.25 m). The temperature program for GC analysis
was set as follows: the temperature was held at 40˚C for 1 min, then raised to 260˚C at
30˚C/min and held for 5 min. Inlet and detector temperatures were 280˚C. The analysis was
carried out directly after sampling to avoid any additional conversion.
2.3 Catalytic test
In a typical catalytic experiment, 0.144 g of the catalyst (corresponding to 0.45 mmol
4
Page 7 of 37

amino groups) was placed in a flask and heated at 150 ˚C for 4 hours. After three cycles of
vacuum pumping and nitrogen injection, the solution of ethyl cyanoacetate or other active
methylene compounds (10 mmol) in an appropriate solvent (5 mL) was added with stirring
and heated to a given temperature. After temperature equilibrium, benzaldehyde (5 mmol)
was added to initiate the reaction. The reaction mixture was stirred under a static nitrogen
atmosphere, and small aliquots of the supernatant were withdrawn at different time intervals
to monitor the reaction conversion by GC.
3 . Results and discussion
3.1 Characterization of the UiO-66-NH2 catalyst.
Fig. 2. PXRD patterns (a) and SEM picture (b) of UiO-66-NH2
UiO-66-NH2 was synthesized according to literature procedures. The XRD pattern of the
5
Page 8 of 37

as-synthesized sample (Fig. 2a) is in good agreement with the literature data and with the
simulated pattern of pristine UiO-66 [24], evidencing good synthetic reproducibility and
phase purity. The SEM morphology (Fig. 2b) shows that the particles are of nanometer size
with a distribution in the range of 40-90 nm. After thermal activation at 150 ˚C, the reflections
at large angles become less intensive, but the main characteristic peaks remain, indicating that
the overall phase integrity is retained although the activation leads to some degree of disorder
in the framework. Fig. 2a also shows the XRD patterns of the UiO-66-NH2 sample after one
and three cycles of catalytic reactions. It is interesting to note that the reflections at large
angles are recovered after the catalytic reaction, indicating that the disorder caused by thermal
treatment is repaired. It can be deduced that the original crystalline phase of the material is
retained without significant degradation in structural integrity after the catalytic reactions.
3.2 Catalytic properties
Table 1. UiO-66-NH2 catalyzed Knoevenagel condensation between benzaldehyde
and
ethyl cyanoacetate: influence of catalytic dose, solvent, and temperature.
Entry
1
Solvent
EtOH
T (˚C)
80
Time (h)
2
Conv (%)
27
2
80
2
67
3
80
2
94
4
5
6
7
8
9
40
40
60
80
40
40
0.5/2
0.5/2
0.5
0.5
0.5/2
2
25/79
27/92
58
91
31/95
5
DMF
DMSO
DCM
10
11
12
EtOAc
THF
Toluene
40
40
40
2
2
2
3
2
<1
Conditions: reactants: benzaldehyde (5 mmol), ethyl cyanoacetate (10 mmol); solvent: 5 ml; Cat:
UiO-66-NH2, 0.144 g (corresponding to 0.45 mmol amino groups), except for entry 1 (no catalyst) and
entry 2 (0.072 g, corresponding to 0.23 mmol amino groups).
6
Page 9 of 37

3.2.1 Influence of catalytic dose, solvent, and temperature
The condensation reaction of benzaldehyde with ethyl cyanoacetate, which affords
-cyanocinnamate, was performed as test reaction to study the effects of different reaction
conditions, including the catalytic dose, the solvent, and the reaction temperature.
The results obtained for the reactions performed in ethanol at 80 ˚C in the presence of
different amount of catalysis are given in Table 1 (entries 1-3). While the blank reaction in the
absence of any catalyst gave a benzaldehyde conversion of only 27% after 2 h, introducing
UiO-66-NH2 and increasing the amount led to significant increase in the conversion. With
0.144 g UiO-66-NH2 (corresponding to 9 mol% amino groups), the conversion reached 94%
after 2 h. For comparison, the blank reaction needs 17 h to reach a conversion of >90%
(complete conversion needs more than 21 h). These results clearly confirm that UiO-66-NH2
is active in catalyzing the Knoevenagel reaction. No by-product was observed during the
reaction, so the selectivity for -cyanocinnamate is 100%.
Strong effects of solvent polarity have been observed for catalytic Knoevenagel reactions.
To establish the solvent effects for UiO-66-NH2, we performed the reaction in different
solvents (Table 1). It proved that the catalyst has the highest activity (> 90 % conversion
within 2 h) in DMSO and DMF (entries 5, 8), which have the highest polarity (dielectric
constants  = 48.9 for DMSO and 36.7 for DMF). Ethanol (protic,  = 24.3, is also a good
medium for the reaction (entry 4). By contrast, nonprotic and less polar solvents, including
DCM ( = 9.1), THF ( = 7.5), EtOAc ( = 6.02) and toluene ( = 2.4), are not good media
(entries 9-12). The effects can be generally explained by the assumption that polar solvents
help to stabilize the charged transition-state complex of the reaction [36]. Besides, the strong
hydrogen-bonding acceptor ability of DMSO and DMF can facilitate proton transfer [45]. On
the one hand, the solvent effects are similar to those observed for some basic organocatalysts
[35, 36] and for the amine-functionalized MOF catalysts IRMOF-3 [45, 46] and
NH2-MIL-101(Al) [48], although sometimes ethanol was found to be better than DMF and
DMSO due to its amphiprotic character. On the other hand, the effects are in clear contrast
with those observed for some amino-tagged silicas, for which toluene is better than polar
solvents [40, 41]. For some other catalysts, the solvent effects are not necessarily related to
the polarity or the amphiprotic nature. For example, the catalytic activity of ZIF-8, -9 and -10
7
Page 10 of 37

(ZIF = zeolite imidazolate framework) for Knoevenagel condensation in toluene was found to
be intermediate between that in THF (higher) and that in DCM (lower) [49, 50]. For
solid-supported base catalysts, it has been reported that the higher the hydrophilicity (or the
surface polarity) of the support, the lower the effect of the solvent [41, 46], but the opposite
trend has also been reported [40]. Therefore, there is no general trend for the effect of solvents
on the Knoevenagel reaction. Nevertheless, the results for UiO-66-NH2, IRMOF-3 and
NH2-MIL-101(Al) consistently indicate that amino-tagged MOFs perform better in highly
polar solvents.
The temperature dependence of the catalytic performance of UiO-66-NH2 was checked by
performing the Knoevenagel reaction in DMF at 40, 60 and 80 ˚C (Table 1, entries 5-7). As
expected, the reaction rate increases quickly with the temperature, the conversion within 30
min changing from 27 % at 40 ˚C to 91 % at 80 ˚C. The conversion at 40 ˚C can reach a high
level (> 90 %) within 2 h.
3.2.2 The influence of different substrates
Having established that UiO-66-NH2 is a good catalyst for the Knoevenagel condensation
between benzaldehyde and ethyl cyanoacetate, we extend the study to various substrates,
including different aromatic aldehydes and methylene compounds. The results are listed in
Table 2.
Diethyl malonate and malononitrile have been compared with ethyl cyanoacetate (entries
1-3). Under the same catalytic conditions, diethyl malonate led to a trace conversion of
benzaldehyde after 2 h, whereas malononitrile caused a conversion of 98% within 40 min.
The different reactivity of the methylene compounds is consistent with their different acidity,
which increases in the order of diethyl malonate (pKa = 16.4 [54]) < ethyl cyanoacetate (pKa =
13.1 [55]) < malononitrile (pKa = 11.1 [56]). Actually, no Ar-NH2-functionalized materials
(including MOFs) have been found to be able to efficiently catalyze the Knoevenagel reaction
between benzaldehyde and diethyl malonate, because the pKa is too high for the amine group
or the in-situ formed imine group to induce deprotonation.
Table 2: UiO-66-NH2 catalyzed Knoevenagel condensation of different substrates.a
8
Page 11 of 37

Entry Aldehyde
1
Methylene
time (min)
120
Conv (%)
92
O
N
O
2
3
4
5
6
120
40
5
< 2
98
100
98
40
40
97
7
40
98
8
40
92
9
40
87
10
11
120
120
56 (21) b
15 (19) b
a Conditions: aldehyde (5 mmol), methylene compound (10 mmol), DMF (5 ml), Catalyst
(0.144 g, corresponding to 0.45 mmol amino groups), 40 ˚C. b The conversion for the control
test in the absence of any catalyst is given in parentheses for comparison.
Various aromatic aldehydes have been tested for catalytic Knoevenagel condensation with
malononitrile. It was found that the activity of UiO-66-NH2 for 2-nitrobenzaldehyde is much
higher than that for benzaldehyde, as the reaction for the former aldehyde can reach
completion within only 5 min (entry 4 in Table 2). This reflects the great accelerating effect of
the strong electron-withdrawing nitro group, as would be expected for a reaction involving
nucleophilic attack at the carbonyl group. The catalyzed reactions of the three
9
Page 12 of 37

methyl-substituted benzaldehydes proceed readily to give  97% conversions after 40 min
(entries 5-7). The conversions are similar to that for benzaldehyde, but the initial reaction
rates of these substrates are different. As shown in Fig. 3, compared with the non-substituted
benzaldehyde, the initial reaction rate of 2-methylbenzaldehyde is similar (or slightly faster),
the reaction of the 3-isomer is obviously slower, and notably, the 4-isomer was more reactive,
though methyl is an electron-donating group. The position effect of methyl substitution on the
reactivity of benzaldehyde is similar to that observed for the Knoevenagel condensation
catalyzed by ZIF-8 [49] and ZIF-9 [50], where the reactivity varies in the order of
4-methylbenzaldehyde  2-methylbenzaldehyde > benzaldehyde > 3-methylbenzaldehyde.
Fig. 3. Time dependence of reaction conversion for the condensation of different
monosubstituted bezaldehydes (5 mmol) with malononitrile (10 mmol) in DMF (5 ml, 40 ˚C)
in the presence of 0.144 g catalyst (corresponding to 0.45 mmol amino groups).
The catalyst system is also efficient for the Knoevenagel reactions of larger aromatic
aldehydes (1-naphthaldehyde and 9-anthraldehyde) with malononitrile, though the conversion
is somewhat decreased when the aldehyde substrate becomes more bulky (comparing entries
3, 8 and 9 in Table 2). The decreased conversion may be a (weak) indicator of size effects.
The effects are more appreciable for the reactions of different aldehydes with ethyl
cyanoacetate (molecular dimensions ~3.4×7.3 Å2), which are bulkier than malononitrile
(~2.8×4.3 Å2). As shown by entries 1 and 10, the conversion of 9-anthraldehyde (56%,
~5.9×9.2 Å) reacting with ethyl cyanoacetate is much lower than that of benzaldehyde (92%,
10
Page 13 of 37

~6.0×4.3 Å2). The results could reflect that the probability of two bulky substrates forming
transition-state complexes is significantly reduced due to the limited space in the catalyst.
To further confirm the size effects, 9-anthraldehyde was reacted with a still bulkier
methylene substrate, tert-butyl cyanoacetate (~4.3×8.2 Å2). The independent control reactions
of 9-anthraldehyde with ethyl and tert-butyl cyanoacetates in the absence of any catalyst give
similar conversions (around 20% after 2 h, entries 10 and 11), indicating that the intrinsic
reactivity of the two methylene substrates is comparable. In the presence of UiO-66-NH2, the
reaction with ethyl cyanoacetate shows a much higher conversion (56%), indicating a
catalytic process. By contrast, the use of UiO-66-NH2 for the reaction with tert-butyl
cyanoacetate leads to a conversion (15%) comparable to (or even slightly lower than) that for
the reaction without any catalyst, indicating that UiO-66-NH2 is not active for the two bulky
substrates. Thus, it is very likely that the catalytic reaction over UiO-66-NH2 proceeds inside
the pores. The pore system of perfect UiO-66 crystals consists of interconnected tetrahedral
and octahedral cages with free diameters of about 8 and 11 Å, respectively. Notably, some
recent studies have demonstrated that real UiO-66 and analogs contains an amount of defects
due to missing linkers between Zr6 clusters [57-59]. The linker vacancies lead to expanded
pores so that the real pore dimensions can be larger. The failure in catalyzing the reaction
between 9-anthraldehyde and tert-butyl cyanoacetate could be because the pores are still too
small for the two bulky substrates diffuse into the pores of the catalyst to access the active
sites and to form the transition state required for the reaction.
3.2.3 Heterogeneity and reusability
With a solid catalyst for liquid-phase reactions, an issue of great concern is whether the
catalytic process is heterogeneous or homogeneous. In the latter case, the species leached into
the liquid phase is partially or fully responsible for the observed catalytic activity. This is
practically undesirable. The heterogeneous nature of our catalyst system for Knoevenagel
reactions is implied in some results of the above catalytic experiments. For example, the
dependence of benzaldehyde conversion on the amount of solid catalyst (Table 1) could be an
implication. The solid shows no appreciable solvability. It means that if there were any
dissolved species, it would be quite easy for the liquid phase to be saturated by the species.
Once saturated, the amount of the species in a given volume of liquid would be independent
11
Page 14 of 37

of the dose of the solid. Then, if the dissolved species were responsible for the catalytic
activity, the reaction conversion would be independent of the amount of the solid. Therefore,
the dependence of conversion on solid dose may be taken as a (weak) indication of
heterogeneous catalysis. A stronger evidence is the size selectivity for substrates (Table 2),
which is difficult to achieve with homogeneous catalysts but characteristic of porous
heterogeneous catalysts. In order to give a more direct confirmation, a control reaction was
performed between benzaldehyde and ethyl cyanoacetate in DMF. After the reaction
proceeded at 40 C for 2 h, the catalyst was removed by hot filtration. The filtrate was further
heated at the same temperature and the composition was monitored at given time intervals.
The results are compared in Fig. 4 with those obtained for the reaction under the same
conditions but without the filtering procedure. Obviously, no appreciable reaction took place
after removal of the catalyst, indicating that no active species were leached into the liquid
phase and providing a direct evidence for the heterogeneity of the catalyst system. This is also
confirmed by ICP-AES analysis, which suggests no detectable Zr in the liquid phase.
Fig. 4. Filtration test of the catalyst. Conditions: benzaldehyde (8 mmol), ethyl cyanoacetate
(7 mmol), Cat. 0.10 g (corresponding to 0.31 mmol amino groups), solvent: DMF (5 ml), 40
˚C. The conversion is based on ethyl cyanoacetate.
Another issue of great concern for solid catalysts is the recyclability. To check this for
UiO-66-NH2, the catalyst was used for three cycles for the reaction between benzaldehyde
and malononitrile. After each cycle, the solid was filtered out, soaked and stirred in methanol
for 24 h, filtered out again and washed with methanol. The recovered solid was heated at 150
12
Page 15 of 37

˚C for 4 hours and then reused in the next cycle. The results of the three cycles are very
similar. As shown in Fig. 5a, the conversion of benzaldehyde after 1 h remains at the high
level of  97%, indicating that the catalyst could be reused without significant degradation in
catalytic performance. The recyclability is consistent with the good chemical resistance of the
structure against the reaction conditions, which has been demonstrated by XRD
measurements (see Fig. 2a). The intactness of the catalyst is also confirmed by the fact the
FT-IR spectra of the catalyst shows no appreciable difference after catalytic reactions (Fig.
5b).
Fig. 5. (a) Catalyst recycling studies. Conditions: benzaldehyde (5 mmol), malononitrile (10
mmol), solvent: DMF (5 ml), Cat. 0.144 g, 40 ˚C, 1 h. (b) IR spectra of the fresh and used
catalysts.
13
Page 16 of 37

3.2.4 Mechanism considerations
Fig. 6. (a) Knoevenagel condensation with different catalysts. Conditions: benzaldehyde (5
mmol), ethyl cyanoacetate(10 mmol), solvent: EtOH (5 ml), 80 ˚C, Cat. 0.45 mmol Zr or NH2,
2 h. (b) The reactions of benzaldehyde (1 mmol) with DMBDC-NH2 (1.2 mmol) in EtOH (6
ml, 80 ˚C) in the absence/presence of UiO-66 (0.014 g).
To determine the role of the amine group in the catalyst, the amine-free MOF (UiO-66)
prepared and activated by similar procedures was tested as catalyst for the reaction between
benzaldehyde and ethyl cyanoacetate. As can be seen from Fig. 6a, under the same catalytic
conditions, the amino-free MOF led to much lower conversion than UiO-66-NH2. Since the
two MOFs are isostructural, the superior performance of UiO-66-NH2 indicates that the basic
amine group is important in promoting the Knoevenagel reaction, as is normally expected for
NH2-tagged catalysts. It has been demonstrated that the introduction of amino groups into the
14
Page 17 of 37

linkers of UiO-66 leads to an appreciable increase in basicity [28]. The organic analogue of
UiO-66-NH2, dimethyl 2-amine-1,4-benzenedicarboxylate (DMBDC-NH2), was also tested
under the same conditions. The homogeneous reaction showed a low conversion of 29 %,
which is only slightly higher than the reaction without any catalysts and much lower than the
heterogeneous reaction using UiO-66-NH2 (see Fig. 6a), though the amount of amino groups
was set to be equal in two catalytic reactions. This could indicate that the catalytic activity of
amine is enhanced in the MOF structure or that the reaction is promoted by other sites besides
amine. The Knoevenagel reactions with aniline as homogenous catalyst have also been
reported to give much lower conversion than those with IRMOF-3 as heterogeneous catalyst
[45, 46]. The high activity of IRMOF-3 compared to aniline was attributed to the increased
basicity of the aromatic amino group when incorporated within the MOF structure. It was
proposed that the increase of basicity could be due to the intramolecular hydrogen bonding
interaction between the amino group and an carboxylate oxygen (coordinated to Zn(II)) [45].
However, this explanation fails for DMBDC-NH2, which could have similar hydrogen
interactions but does not show significant catalytic activity. A DFT study [60] suggested that
the basicity of the aromatic amine (Ar-NH2) and the aldimine derivative (Ar-N=CHC6H5)
indeed increases in going from aniline to DMBDC-NH2 and that the increase is more
significant in going further to the amine embedded in IRMOF-3. The increase has been
related to the stabilization of the protonated aminium and iminium species, which are
hydrogen bonded to a carboxylate oxygen to generate a 6-membered planar ring. The
theoretical study also indicated that the aldimine intermediates are the most basic species in
the reaction systems and should be the active species that deprotonate ethyl cyanoacetate.
However, the DFT energetic study on the catalytic cycles of Knoevenagel condensation
indicated that the increased catalytic activity of IRMOF-3 seems to be unrelated to its
increased basicity but could be explained by its adsorption ability for the water by-product
which would otherwise poisons the catalytic amino sites.
It should be noted that the theoretical study was based on the hypothesis that the amine
groups are the only catalytic sites in the MOF, not considering the possibility of the existence
of other active sites. A recent reinvestigation of IRMOF-3 for Knoevenagel condensation
suggested that the MOF behaves as an 'unexpected' bifunctional acid-base catalyst [15]. The
15
Page 18 of 37

'unintentional' acid sites come from a defective origin, either Zn–OH species formed upon
partial hydrolysis of the framework or ZnO particles entrapped inside the cavities during the
synthesis.
As already mentioned, UiO-66 is much less active than UiO-66-NH2, but it is noteworthy
that the NH2-free MOF still show significant catalytic activity (50% after 2 h, See Fig. 6a)
when compared with the blank non-catalytic reaction. This suggests that the Knoevenagel
reaction could also be promoted by Zr(IV) centers (Lewis acid sites), so UiO-66-NH2 could
be a bifunctional acid-base catalyst for Knoevenagel condensation. The bifunctional character
has recently been proposed to explain the high activity of the material for the cross-aldol
condensation between heptanal and benzaldehyde [24]. The acid sites in UiO-66 and
UiO-66-NH2 could be the open Zr(IV) sites generated by dehydration of the Zr6O4(OH)4
cluster to give Zr6O6. It has been shown that the dehydration begins at about 100 ˚C for
UiO-66 and even below the temperature for UiO-66-NH2 [24]. Therefore, the activation
procedure (heating at 150 ˚C) that we applied before catalytic reactions could induce partial
dehydration. Another origin of the Lewis acid site could be the presence of defects. Since the
UiO-66 MOFs are much stable than MOF-5 and IRMOF-3, one would not expected that the
Zr MOFs contains a large density of defects of similar origin to those in the Zn MOFs.
However, recent studies have demonstrated the presence of linker vacancies (defects due to
the missing of organic linkers between Zr6 clusters) even in well-crystallized UiO-66
materials [57-59]. The defects could generate open metal sites upon heating. Whatever the
origin of the acid sites is, the presence of both acid and base sites could explain the superior
performance of UiO-66-NH2.
Comparisons among ZrO2, DMBDC-NH2 and their mixture may give some information
about the acid-base cooperation. As shown in Fig. 6a, ZrO2 and DMBDC-NH2 show very low
activity (the conversions are higher than the blank test by only 8% and negligibly 2%,
respectively), while the mixture show enhanced activity (15% higher than the blank test),
indicating the presence of some cooperation, although weak. The much higher activity of
UiO-66-NH2 than UiO-66 may indicate much stronger cooperation in the porous framework.
Probably, the framework arranges Zr and amino sites in close proximity on the inner surface
so that the two sites can cooperate more efficiently.
16
Page 19 of 37

Knoevenagel condensation, as a modification of aldol condensation, may proceed via
different mechanisms that depend on the nature of the catalyst used. A Lewis acid site usually
interacts with the carbonyl oxygen of benzaldehyde, thus the C=O bond is further polarized to
facilitate the nucleophilic attack at the carbon atom. Strong bases would cause direct
deprotonation of the methylene group, generating the carbanion that attacks the carbonyl
carbon atom of benzaldehyde. For amino-based catalysts, it is generally accepted that the
reaction proceeds via aldimine intermediates [45, 60]. Two nucleophilic addition-elimination
processes are involved, as illustrated in Scheme 2a. The first is the formation of an aldimine
intermediate (and water) from aldehyde and the amine group at the catalyst surface. The
aldimine intermediate is more basic than the original amine and also more reactive than the
original aldehyde [36]. Thus, the methylene compound is activated (deprotonated) by
aldimine rather than by amine, and meanwhile the aldimine group is also activated by the
proton transfer. Therefore, the second addition-elimination process is facilitated, which occurs
between methylene and aldimine (rather than aldehyde) groups to give the product and to
regenerate the amine catalyst. Spectroscopic evidences for the formation of intermediate
aldimine have been recently reported for amino-tagged silicas and IRMOF-3 [45, 61].
According to this mechanism, the formation of the aldimine intermediate is important.
We have studied the formation by reacting benzaldehyde with DMBDC-NH2 under the
conditions used for the Knoevenagel reaction. As can be seen from Fig. 6b, the formation of
aldimine in the absence of any catalyst is much slower than in the presence of UiO-66. The
results clearly suggest the catalytic role of the Zr MOF for aldimine formation. The promotion
is most likely due to the activation of carbonyl groups by Zr sites.
Based on the above considerations, the cooperative acid-base bifunctional catalysis of
UiO-66-NH2 can be speculated as follows (Scheme 2b). The Zr site serves to activate
benzaldehyde so that the amino group in close proximity the Zr site can readily attack the
carbonyl group to generate the aldimine intermediate, which serves not only as a base (the N
atom) to extract proton from methylene but also as an acid (the C atom) to couple with
methylene. This synergic process is a little different from that proposed for the cross-aldol
condensation between heptanal and benzaldehyde [24], where the reaction was supposed to
occur between benzaldehyde activated by Zr(IV) and the heptanal methylene group activated
17
Page 20 of 37

by amine.
Scheme 2. (a) Mechanism for amine-catalyzed Knoevenagel condensation. (b) Aldimine
formation promoted by the Lewis acid site in close proximity to the amine group.
4. Conclusions
In this work, we reported the use of UiO-66-NH2 as solid catalyst for Knoevenagel
condensation. It has been demonstrated that the catalyst shows better performance in highly
polar solvents such as DMF, DMSO and ethanol. It can efficiently catalyze the condensation
reactions of aromatic aldehydes with cyanoacetate and malononitrile, the conversion easily
reaching more than 90% under mild conditions within a short time (dependent on the
substrates). The catalytic process is heterogeneous and shows size effects, characteristic of a
porous catalyst with active sites inside the pores. The catalyst can be recycled without losing
its framework integrity and catalytic activity. Compared with the homogeneous amino
precursor (2-aminoterephthalate) and the isostructural amino-free MOF (UiO-66), the
superior performance of UiO-66-NH2 has been attributed to the acid-base bifunctional
18
Page 21 of 37

character. It was proposed that the Zr site in close proximity to the amino group facilitates the
formation of aldimine intermediates, which are the active species reacting with the methylene
compounds. This study further demonstrates that MOFs can offer extraordinary opportunity of
achieving site-isolated acid-base catalysts, which are desirable for more efficient and more
facile organic synthesis. Working along this line, we are exploring such catalysts for
multi-component reactions.
Acknowledgements
This work is supported by the National Science Foundation of China (NSFC nos. 21173083
and 91022017) and the Research Fund for the Doctoral Program of Higher Education of
China.
References
[1] C. Wang, D. Liu, W. Lin, J. Am. Chem. Soc. 135 (2013) 13222-13234.
[2] H. Furukawa, K. E. Cordova, M. O'Keeffe, O. M. Yaghi, Science 341 (2013)
1230444(1-12).
[3] M. P. Suh, H. J. Park, T. K. Prasad, D. W. Lim, Chem. Rev. 112 (2012) 782-835.
[4] K. Sumida, D. L. Rogow, J. A. Mason, T. M. McDonald, E. D. Bloch, Z. R. Herm, T. H.
Bae, J. R. Long, Chem. Rev. 112 (2012) 724-781.
[5] S. Xiang, Y. He, Z. Zhang, H. Wu, W. Zhou, R. Krishna, B. Chen, Nat. Commun. 3 (2012)
1956(1-9).
[6] J.-R. Li, Y.-G. Ma, M.C. McCarthy, J. Sculley, J.-M. Yu, H.-K. Jeong, P.B. Balbuena, H.-C.
Zhou, Coord. Chem. Rev. 255 (2011) 1791-1823.
[7] Z. Zhang, Y. Zhao, Q. Gong, Z. Li, J. Li, Chem. Commun. 49 (2013) 653-661.
[8] M. Yoon, R. Srirambalaji, K. Kim, Chem. Rev. 112 (2012) 1196-1231.
[9] P. Valvekens, F. Vermoortele, V. D. De, Catal. Sci. Technol. 3 (2013) 1435-1445.
[10] A. Dhakshinamoorthy, M. Opanasenko, J. Cejka, H. Garcia, Catal. Sci. Technol. 3 (2013)
2509-2540.
[11] A. Corma, H. Garcia, F. X. L. Xamena, Chem. Rev. 110 (2010) 4606-4655.
[12] B. Voit, Angew. Chem. Int. Ed. 45 (2006) 4238-4240.
19
Page 22 of 37

[13] N. R. Shiju, A. H. Alberts, S. Khalid, D. R. Brown, G. Rothenberg, Angew. Chem. Int.
Ed. 50 (2011) 9615-9619.
[14] R. Srirambalaji, S. Hong, R. Natarajan, M. Yoon, R. Hota, Y. Kim, Y. Ho Ko, K. Kim,
Chem. Commun. 48 (2012) 11650-11652.
[15] F. X. Llabrés i Xamena, F. G. Cirujano, A. Corma, Microporous and Mesoporous Mater.
157 (2012) 112-117.
[16] A. Dhakshinamoorthy, M. Alvaro, A. Corma, H. Garcia, Dalton Trans. 40 (2011)
6344-6360.
[17] S. M. Cohen, Chem. Rev. 112 (2012) 970-1000.
[18] J. Juan-Alcaniz, J. Ferrando-Soria, I. Luz, P. Serra-Crespo, E. Skupien, V.P. Santos, E.
Pardo, i.X.F.X. Llabres, F. Kapteijn, J. Gascon, J. Catal. 307 (2013) 295.
[19] Y. B Huang, S. Y. Gao, T. F. Liu, J. Lü, X. Lin, H. F. Li, R. Cao, Chem. Plus. Chem. 77
(2012) 106–112.
[20] Y. B. Huang, T. Ma, P. Huang, D. S. Wu, Z. J. Lin, R. Cao, Chem. Cat. Chem. 5 (2013)
1877–1883.
[21] Y.B. Huang, Z. J. Lin, R. Cao, Chem. Eur. J. 17 (2011) 12706-12712.
[22] Y. B. Huang, S. J. Liu, Z. J. Lin, W. J. Li, X. F. Li, R. Cao, J. Catal. 292 (2012) 111.
[23] J. H. Cavka, S. Jakobsen, U. Olsbye, N. Guillou, C. Lamberti, S. Bordiga, K. P. Lillerud,
J. Am. Chem. Soc. 130 (2008) 13850-13851.
[24] F. Vermoortele, R. Ameloot, A. Vimont, C. Serre, D. De Vos, Chem. Commun. 47 (2011)
1521-1523.
[25] F. Vermoortele, M. Vandichel, D. V. B. Van, R. Ameloot, M. Waroquier, S.V. Van, V. D. E.
De, Angew. Chem. Int. Ed. 51 (2012) 4887-4890.
[26] F. Vermoortele, B. Bueken, B. G. Le, D. V. B. Van, M. Vandichel, K. Houthoofd, A.
Vimont, M. Daturi, M. Waroquier, S. V. Van, C. Kirschhock, V. D. E. De, J. Am. Chem. Soc.
135 (2013) 11465-11468.
[27] J. Kim, S.-N. Kim, H.-G. Jang, G. Seo, W.-S. Ahn, Appl. Catal., A. 453 (2013) 175-180.
[28] M. N. Timofeeva, V. N. Panchenko, J. W. Jun, Z. Hasan, M. M. Matrosova, S. H. Jhung,
Appl. Catal., A. 471 (2014) 91-97.
[29] D. Sun, Y. Fu, W. Liu, L. Ye, D. Wang, L. Yang, X. Fu, Z. Li, Chem. Eur. J. 19 (2013)
20
Page 23 of 37

14279-14285.
[30] L. Shen, W. Wu, R. Liang, R. Lin, L. Wu, Nanoscale 5 (2013) 9374-9382.
[31] L. Shen, S. Liang, W. Wu, R. Liang, L. Wu, Dalton Trans. 42 (2013) 13649-13657.
[32] J. Long, S. Wang, Z. Ding, S. Wang, Y. Zhou, L. Huang, X. Wang, Chem. Commun. 48
(2012) 11656-11658.
[33] C. Wang, Z. Xie, K. K. E. de, W. Lin, J. Am. Chem. Soc. 133 (2011) 13445-13454.
[34] C. G. Silva, I. Luz, i. X. F. X. Llabres, A. Corma, H. Garcia, Chem. Eur. J. 16 (2010)
11133-11138.
[35] M. Climent, A. Corma, I. Dominguez, S. Iborra, M. Sabater, G. Sastre, J. Catal. 246
(2007) 136-146.
[36] I. Rodriguez, G. Sastre, A. Corma, S. Iborra, J. Catal. 183 (1999) 14-23.
[37] S. K. Mohamed, A. A. Abdelhamid, A. M. Maharramov, A. N. Khalilov, A. V. Gurbanov,
M. A. Allahverdiev, J. Chem. Pharm. 4 (2012) 1787-1793
[38] G. Bartoli, M. Bosco, A. Carlone, R. Dalpozzo, P. Galzerano, P. Melchiorre, L. Sambri,
Tetrahedron Lett. 49 (2008) 2555-2557.
[39] B. M. Choudary, M. Lakshmi Kantam, V. Neeraja, K. Koteswara Rao, F. Figueras, L.
Delmotte, Green Chem. 3 (2001) 257-260.
[40] D. J. Macquarrie, T. Yoshitomi, K. Seki, Chem. Commun. (1997) 1781-1782.
[41] A. Corma, S. Iborra, I. Rodriguez, F. Sanchez, J. Catal. 211 (2002) 208-215.
[42] S. Hasegawa, S. Horike, R. Matsuda, S. Furukawa, K. Mochizuki, Y. Kinoshita, S.
Kitagawa, J. Am. Chem. Soc. 129 (2007) 2607-2614.
[43] Y. K. Hwang, D.-Y. Hong, J.-S. Chang, S. H. Jhung, Y.-K. Seo, J. Kim, A. Vimont, M.
Daturi, C. Serre, G. Ferey, Angew. Chem. Int. Ed. 47 (2008) 4144-4148.
[44] S. Neogi, M. K. Sharma, P. K. Bharadwaj, J. Mol. Catal. A: Chem. 299 (2009) 1-4.
[45] J. Gascon, U. Aktay, M. Hernandezalonso, G. Vanklink, F. Kapteijn, J. Catal. 261 (2009)
75-87.
[46] A. R. Burgoyne, R. Meijboom, Chem. Lett. 143 (2013) 563-571.
[47] M. Opanasenko, A. Dhakshinamoorthy, M. Shamzhy, P. Nachtigall, M. Horacek, H.
Garcia, J. Cejka, Catal. Sci. Technol. 3 (2013) 500-507.
[48] P. Serra-Crespo, E. V. Ramos-Fernandez, J. Gascon, F. Kapteijn, Chem. Mater. 23 (2011)
21
Page 24 of 37

2565-2572.
[49] U. P. N. Tran, K. K. A. Le, N. T. S. Phan, ACS Catal. 1 (2011) 120-127.
[50] L. T. L. Nguyen, K. K. A. Le, H. X. Truong, N. T. S. Phan, Catal. Sci. Technol. 2 (2012)
521.
[51] S.-N. Kim, S.-T. Yang, J. Kim, J.-E. Park, W.-S. Ahn, Cryst. Eng. Comm. 14 (2012)
4142.
[52] A. Schaate, P. Roy, A. Godt, J. Lippke, F. Waltz, M. Wiebcke, P. Behrens, Chem. Eur. J.
17 (2011) 6643.
[53] S. J. Garibay, S. M. Cohen, Chem. Commun., 46 (2010) 7700-7702.
[54] B. R. Gelin, M. Karplus, J. Am. Chem. Soc. 26 (1975) 6996-7006.
[55] F. G. Bordwell, H. E. Fried, J. Org. Chem. 46 (1981) 4328-4331.
[56] W. N. Olmstead, F. G. Bordwell, J. Org. Chem. 45 (1980) 3299-3305
[57] H. Wu, Y. S. Chua, V. Krungleviciute, M. Tyagi, P. Chen, T. Yildirim, W. Zhou, J. Am.
Chem. Soc. 135 (2013) 10525.
[58] L. Valenzano, B. Civalleri, S. Chavan, S. Bordiga, M. H. Nilsen, S. Jakobsen, K. P.
Lillerud, C. Lamberti, Chem. Mater. 23 (2011) 1700-1718.
[59] F. Vermoortele, B. Bueken, G. Le Bars, B. Van de Voorde, M. Vandichel, K. Houthoofd,
A. Vimont, M. Daturi, M. Waroquier, V. Van Speybroeck, C. Kirschhock, D. E. De Vos, J. Am.
Chem. Soc. 135 (2013) 11465-11468.
[60] R. Cortese, D. Duca, Phys. Chem. Chem. Phys. 13 (2011) 15995-16004.
[61] R. Wirz, D. Ferri, A. Baiker, Langmuir 22 (2006) 3698-3706
22
Page 25 of 37

Captions for Schemes and Figures
Scheme 1. Knoevenagel condensation.
Scheme 2. (a) Mechanism for amine-catalyzed Knoevenagel condensation. (b) Aldimine
formation promoted by the Lewis acid site in close proximity to the amine group.
Fig. 1. Structure of UiO-66. (a) A [Zr6(µ3-O)4(µ3-OH)4(µ2-COO)12] cluster. (b) The porous
systems in which octahedral and tetrahedral cages share triangular windows.
Fig. 2. PXRD patterns (a) and SEM picture (b) of UiO-66-NH2
Fig. 3. Time dependence of reaction conversion for the condensation of different
monosubstituted bezaldehydes (5 mmol) with malononitrile (10 mmol) in DMF (5 ml, 40 ˚C)
in the presence of 0.144 g catalyst (corresponding to 0.45 mmol amino groups).
Fig. 4. Filtration test of the catalyst. Conditions: benzaldehyde (8 mmol), ethyl cyanoacetate
(7 mmol), Cat. 0.10 g (corresponding to 0.31 mmol amino groups), solvent: DMF (5 ml), 40
˚C. The conversion is based on ethyl cyanoacetate.
Fig. 5. (a) Catalyst recycling studies. Conditions: benzaldehyde (5 mmol), malononitrile (10
mmol), solvent: DMF (5 ml), Cat. 0.144 g, 40 ˚C, 1 h. (b) IR spectra of the fresh and used
catalysts.
Fig. 6. (a) Knoevenagel condensation with different catalysts. Conditions: benzaldehyde (5
mmol), ethyl cyanoacetate(10 mmol), solvent: EtOH (5 ml), 80 ˚C, Cat. 0.45 mmol Zr or NH2,
2 h. (b) The reactions of benzaldehyde (1 mmol) with DMBDC-NH2 (1.2 mmol) in EtOH (6
ml, 80 ˚C) in the absence/presence of UiO-66 (0.014 g).
23
Page 26 of 37

Table 1. UiO-66-NH2 catalyzed Knoevenagel condensation between benzaldehyde
and
ethyl cyanoacetate: influence of catalytic dose, solvent, and temperature.
Entry
1
Solvent
EtOH
T (˚C)
80
Time (h)
2
Conv (%)
27
2
80
2
67
3
80
2
94
4
5
6
7
8
9
40
40
60
80
40
40
0.5/2
0.5/2
0.5
0.5
0.5/2
2
25/79
27/92
58
91
31/95
5
DMF
DMSO
DCM
10
11
12
EtOAc
THF
Toluene
40
40
40
2
2
2
3
2
<1
Conditions: reactants: benzaldehyde (5 mmol), ethyl cyanoacetate (10 mmol); solvent: 5 ml; Cat:
UiO-66-NH2, 0.144 g (corresponding to 0.45 mmol amino groups), except for entry 1 (no catalyst) and
entry 2 (0.072 g, corresponding to 0.23 mmol amino groups).
24
Page 27 of 37

Table 2: UiO-66-NH2 catalyzed Knoevenagel condensation of different substrates.a
Entry Aldehyde
Methylene
time (min)
Conv (%)
O
N
1
120
92
O
2
3
4
5
6
120
40
5
< 2
98
100
98
40
40
97
7
40
98
8
40
92
9
40
87
10
11
120
120
56 (21) b
15 (19) b
a Conditions: aldehyde (5 mmol), methylene compound (10 mmol), DMF (5 ml), Catalyst
(0.144 g, corresponding to 0.45 mmol amino groups), 40 ˚C. b The conversion for the control
test in the absence of any catalyst is given in parentheses for comparison.
25
Page 28 of 37

Scheme 1
26
Page 29 of 37