Knoevenagel condensation reactions catalysed by metal-organic frameworks

Article abstract of DOI:10.1007/s10562-013-0995-5

Functionalised metal-organic frameworks (MOFs) which contain amino groups on their secondary building units (SBUs) are basic catalysts in the Knoevenagel condensation. University of Michigan Crystalline Material-1-amine (UMCM-1-NH₂), Isoreticul

Full text of DOI:10.1007/s10562-013-0995-5

Catal Lett (2013) 143:563–571
DOI 10.1007/s10562-013-0995-5
Knoevenagel Condensation Reactions Catalysed
by Metal-Organic Frameworks
Andrew R. Burgoyne Reinout Meijboom
•
Received: 6 February 2013 / Accepted: 17 March 2013 / Published online: 30 March 2013
Springer Science+Business Media New York 2013
ꢀ
Abstract Functionalised metal-organic frameworks (MOFs)
which contain amino groups on their secondary building
units (SBUs) are basic catalysts in the Knoevenagel con-
densation. University of Michigan Crystalline Material-1-
and addition reactions are some of the more important
reaction steps employed for building large and complex
molecules, characteristic of the fine chemical and phar-
maceutical industries. The replacement of liquid organic
bases such as piperidine, as well as alkali metal hydroxides,
with solid base catalysts could lead to an easier separation
of catalyst from the product stream, and thus to a cleaner
production. In addition, heterogeneous catalysts are known
to suppress side reactions such as self-condensation and
oligomerization, resulting in better selectivity and reaction
conversion.
amine (UMCM-1-NH ), Isoreticular MOF-3 (IRMOF-3)
2
and a 5 % IRMOF-3 mixed in Isoreticular MOF-1 (MI-
XMOF) were evaluated as solid basic catalysts and dem-
onstrated to be more active than aniline, their homogenous
anologue. UMCM-1-NH ’s performance was higher than
2
that of the other MOFs studied. IRMOF-3 and MIXMOF
proved to have equal catalytic ability. Their catalytic per-
formance was monitored at various temperatures and in
different solvents for the reaction between benzaldehyde
and ethyl acetoacetate or ethyl cyanoacetate.
The Knoevenagel condensation occurs between a C=O
group and an activated methylene group. This C–C bond
coupling reaction is useful within the drug industry where
important intermediates are prepared. In addition, it is
employed as a classic test reaction to analyse the activity of
a variety of solid base catalysts. The hydrogens in
Keywords Heterogeneous catalysis ꢀ Catalysis activity ꢀ
Elementary kinetics ꢀ C–C coupling ꢀ Organic chemicals
and reactions
b-dicarbonyl compounds have typical pK values in the
a
range 9–11 [4]. This makes them ideal substrates for test-
reactions utilizing a Knoevenagel condensation. Com-
pounds with active methylene groups, with different pKa
values can be reacted with belzaldehyde to test the
strength of a basic catalyst. A robust study using IRMOF-3
1
Introduction
The potential use of porous base catalysts in fine chemical
production is enormous [1–3]. Base-catalysed condensation
was reported previously using ethyl cyanoacetate (pK *9)
a
and ethyl acetoacetate (pK *10.7) with benzaldehyde
a
[
5].
Here we report on the Knoevenagel condensation
between benzaldehyde and ethyl cyanoacetate or ethyl
acetoacetate as catalysed by the primary amines embedded
inside the well-known metal-organic framework IRMOF-3
Electronic supplementary material The online version of this
article (doi:10.1007/s10562-013-0995-5) contains supplementary
material, which is available to authorized users.
[
6]. Initial catalyst evaluation resulted in a set of optimized
A. R. Burgoyne ꢀ R. Meijboom (&)
conditions that was expanded to other MOF frameworks,
Research Centre for Synthesis and Catalysis, Department of
Chemistry, University of Johannesburg, PO Box 524, Auckland
Park, Johannesburg 2006, South Africa
namely MIXMOF [7] and UMCM-1-NH [8]. All reactions
2
were compared against the homogeneous counterpart uti-
lizing aniline as the catalyst.
e-mail: rmeijboom@uj.ac.za
123

5
64
A. R. Burgoyne, R. Meijboom
2
Experimental
reaction often used to test catalytic abilities of basic cata-
lysts [9] and in the production of fine chemicals as well as
heterocyclic compounds of biological significance [10].
Conventionally, this reaction is catalysed by alkali metal
hydroxides or by homogeneous organic bases like primary,
secondary and tertiary amines [11]. However, recycling
and catalyst recovery are problems which arise from their
use.
2
.1 General Experimental
All chemicals were purchased from Sigma-Aldrich and used
as received. GC runs were performed on a Shimadzu
GC-2010-Plus with an auto injector (AOC-20i), auto sam-
ꢁ
pler (AOC-20s) and a 30 m RTX -5 fused silica capillary
column. The synthesis of IRMOF-3 [6], MIXMOF [7] and
UMCM-NH2 [8] was performed according to literature
procedures. All characterisation data were in agreement with
literature (see supplementary material).
The amine catalysed Knoevenagel condensation reac-
tion, Scheme 1, involves the nucleophilic attack of the
amine onto the methylene substrates resulting in a carbo-
cation intermediate. The carbo-cation intermediate then is
involved in a nucleophilic addition with benzaldehyde
resulting in the formation of the product and water with
2
.2 Catalytic Runs
100 % selectivity and regeneration of the amine catalyst.
All catalytic runs were performed in a similar manner. A
representative example is given here. A given amount of the
amino containing catalyst, corresponding to 0.2 mmol –NH2,
was added to a solution of ethyl acetoacetate or ethyl cya-
The metal-organic frameworks IRMOF-3, UMCM-1-
NH and MIXMOF were tested as potential basic catalysts
2
in the Knoevenagel condensation reaction between benz-
aldehyde and different methylene substrates in a range of
solvents at different temperatures. Initial tests were per-
3
noacetate (7 mmol) in solvent (5 cm ) in a round bottom flask
3
50 cm ). The reactor was encapsulated in a water jacket for
(
formed using diethyl malonate which has a pK of 13.3 and
a
temperature control, while being shaken under an inert atmo-
benzaldehyde utilizing IRMOF-3 as the catalyst. The pKa
of this substrate, however, proved to be too high and only
1.6 ± 0.9 % of the diethyl malonate was converted to the
desired product after 130 min in DMSO at 318 K. The
reaction was allowed to continue up to 20 h where it
revealed that only 5.9 ± 1.1 % conversion was achieved
with 100 % selectivity.
sphere (N ) to avoid oxidation. After temperature equilibrium
2
was reached, benzaldehyde was added (8 mmol) while keep-
ing the reaction mixture under a static nitrogen atmosphere.
The reaction mixture was sampled at predetermined time-
3
intervals. A 40 ll sample was diluted with CH Cl (1.9 cm )
2
2
and of this mixture 0.5 ll was injected into the GC. The
analysis was performed immediately after sampling to avoid
any additional conversion in the reaction mixture.
Subsequently, ethyl acetoacetate was utilized as a sub-
strate, with a pK of 10.7. The lowered pK resulted in
a
a
increased conversion. In the absence of a catalyst, some
product formed (3.9 ± 1.8 %), similar to the reaction using
diethyl malonate as a substrate (1.6 ± 0.9 %). A compar-
ison was thus performed to analyse the performance of
IRMOF-3, a porous heterogenous catalyst, against aniline,
a homogenous analogue similar to that found within the
MOF structure.
2
.3 Computational Details
The averages of triplicate catalytic runs were modelled onto
(
a line of best fit using the equation y = -A e
-Rx)
? c using
Kinetic Studio. The gradient of these exponential slopes
provided the kobs for each catalytic run with a corresponding
standard deviation. Manipulation of these provided infor-
mation to calculate the activation energy from the Arrhenius
equation and the enthalpy and entropy from the Eyring Plot.
An ANOVA statistical analysis, using Microsoft Excel 2010,
was performed on the final conversion of the homogenous vs
heterogeneous, solvent and temperature studies to statisti-
cally analyse their differences. A Student’s t test was per-
formed on the final conversion of the MeOH/EtOH at 353
and 333 K, the order of addition of substrates and different
substrates studies to compare two different variables.
The products of the Knoevenagel condensation between
benzaldehyde and ethyl acetoacetate were characterized
using conventional spectroscopic techniques. Filtration, to
remove the IRMOF-3 catalyst, and subsequent removal of
the solvent in vacuo resulted in a clean product. Conven-
tional FT-IR and NMR analyses on the yellow oil obtained
confirmed the presence of the proposed condensation
product.
FT-IR analysis of the substrates, benzaldehyde and ethyl
acetoacetate, and product gave vital information about the
recognition of their disappearance of substrates and the
appearance of the product. The product showed identifiable
-
1
3
Results and Discussion
absorption peaks at mC=C = 1,624 cm
formed C=C bond; mC=O = 1,727 and 1,705 cm for the
b-dicarbonyl system which shifted from m = 1,737 and
for the newly
-
1
The Knoevenagel condensation, of aldehydes with com-
pounds containing activated methylene groups, is a robust
C=O
-
1
-1
1,713 cm
in ethyl acetoacetate; m=C–H = 2,995 cm
1
23

Knoevenagel Condensation Reactions
565
Scheme 1 Catalytic cycle for
Knoevenagel condensation
between benzaldehyde and
i ethyl acetoacetate or ii ethyl
cyanoacetate, catalysed by a
primary amine as contained in
the MOF materials
O
O
O
(
(
i)
O
O
+ H O
2
O
O
O
(i)
ii)
NH2
O
O
(ii)
CN
CN
O
O
O
O
O
(
(
i)
(i)
NH3
O
O
O
(ii)
O
O
CN
ii)
O
H
O
O
CN
-
1
and d=C–H = 696 cm for the newly formed proton on the
initial data points (10 min into reaction) and based on the
-1
alkene; m
1
= 1,044 cm which shifted from mC–O–C
,040 cm in ethyl acetoacetate. The product showed an
=
assumption that all –NH groups are catalytically active cen-
2
C–O–C
-1
tres, no matter how deep within the porous MOF network.
-1
absence of peaks in the region 2,720–2,850 cm typical for
the methylene hydrogens in ethyl acetoacetate, m
=
C–H
3.1 Optimization of Reaction Conditions
=
-1
2
,984 cm . Also absent in the product were m
= 2,818
C–H
=
-1
-1
and 2,737 cm and mC=O = 1,698 cm , which represent
the benzaldehyde’s carbonyl hydrogen and the carbonyl.
A range of conditions could potentially influence the
Knoevenagel condensation reaction. Thus, in order to be
able to compare the various potential catalysts, we opti-
mized the conditions for IRMOF-3. Subsequently, the
different MOF materials were compared under optimized
conditions. The following parameters were studied: solvent
dependence, temperature and order of addition.
1
H NMR analysis of the reaction mixture was performed
at the start of a reaction and then again upon completion of
the 130 min. The singlet at 10.13 ppm (CHO) for benzal-
dehyde was of lower intensity than that of the peaks of the
product. New absorbance peaks at 1.29 ppm (t, 3H, CH3,
J = 6.8 Hz), 2.39 ppm (s, 3H, CH ), 4.28 ppm (q, 2H,
3
A solvent dependence study was performed at 353 K to
determine which factors would influence the activity of the
catalysts. The reaction rate of a homogenous basic catalyst
depends on several aspects, including:
CH , J = 7.2 Hz), 7.34–7.40 ppm (m, 3H, CH), 8.01 ppm
2
(
d, 2H, CH) and 8.40 ppm (s, 1H, C=CH) were present,
confirming the presence of the product.
The heterogeneous catalyst, IRMOF-3, gave significantly
higher conversion rates in DMSO at 353 K and was faster,
•
•
Polarity: when polar reagents are involved, the transi-
tion-state complex is better able to be solvated by polar
solvent molecules. This behaviour of the solvent influ-
encing the transition state is accounted for in the ability of
the catalyst to transfer protons. The general ruleis that the
more polar solvents show higher reaction rates [12, 13].
Amphiprotic properties: a protic solvent (i.e. ethanol or
methanol)might alsoincrease the activation of the slightly
basic benzaldehyde, resulting in higher activities.
-2 -1
kobs = (6.71 ± 1.21) 910
equivalent, aniline which had a k = (0.30 ± 0.01)
s , than the homogenous
obs
-
2 -1
9
10
equal in both cases (Table 1). Aniline closely resembles the
-aminoterephthalic acid which contains the amino group
s , even with the amount of amino groups being
2
within the MOF material. Gascon et al. [5] also observed a
similar trend when working with ethyl cyanoacetate in
DMSO. This indicates that the reactivity of the amino groups
is increased by being contained within the MOF structure. In
the absence of a catalyst only 3.9 ± 1.8 % of the substrates
The solvents selected, for solvent dependence study of
the MOF catalysed reaction between benzaldehyde and
ethyl acetoacetate, were solvents with a range of dielectric
constants (e): DMSO (e = 48.9), DMF (e = 36.7), MeOH
-
2 -1
were converted to product, kobs = (0.00 ± 0.17) 910
The turn-over-frequencies (TOFs) given are based on the
s .
123

5
66
A. R. Burgoyne, R. Meijboom
Table 1 Reaction conversions, calculated kobs values and TOFs for solvents studied in Knoevenagel condensation of benzaldehyde and ethyl
acetoacetate catalysed by IRMOF-3, aniline and without catalyst in DMSO at 353 K
-
2
-1
-1
Initial TOF (min )
Catalyst
Solvent
DMSO
Temperature (K)
353
Reaction conversion (%)
k
obs (10
s
)
IRMOF-3
Aniline
64.5 ± 4.4
28.0 ± 3.6
3.9 ± 1.8
67.08 ± 12.08
3.00 ± 0.99
0.00 ± 1.65
0.44
0.02
–
No catalyst
Fig. 1 Knoevenagel
condensation of benzaldehyde
and ethyl acetoacetate in DMSO
(
filled diamond); DMF (filled
square); MeOH (filled triangle);
EtOH (white circle) or toluene
(times) at 353 K catalysed by
IRMOF-3; n = 3 repeats
(
e = 33), EtOH (e = 24.3) and toluene (e = 2.4). Solvents
with higher dielectric constants were assumed to have a
Zn atoms in IRMOF-3 are fully coordinated [9], i.e. have a
full coordination sphere, resulting in IRMOF-3 being a
hydrophobic catalyst support. Hence, a large effect on the
catalytic ability of many Zn based MOFs is due to the effect
of solvents. However, it was noted that at 353 K in the more
polar solvents the amount of catalyst able to be recycled, by
filtration and washing with fresh DMF, was significantly less
than in solvents with lower polarities. Nguyen et al. [14]
showed that zeolite imidazolate frameworks (ZIFs), a novel
subclass of MOFs, also suffers from a decrease in the crystal
size, in THF (e = 7.6), due to the crystals breaking into small
fragments. Thus, a sacrificial balance between activity and
recyclability needs to be made.
higher polarity.
The results for the solvent dependence study are shown
in Fig. 1 where benzaldehyde was reacted with ethyl ace-
3
toacetate at 353 K in 5 cm of different solvents with IR-
MOF-3 as a catalyst. DMSO, with the highest dielectric
constant, had the highest conversion of 64.5 % ± 4.4 with
-
kobs = (10.53 ± 0.52) 910
by the hydrogen bond acceptor power of DMSO, which
2 -1
s . This can be explained
stabilises the –NH group in the protonated form, and thus
2
increases the rate of proton transfer [12] from the methylene
substrate to the amine catalyst. As a result, the intermediate
is stabilized by the more polar solvents. In all runs, typical
first order kinetic plots were obtained, as shown in Fig. 1.
The calculated rates of reaction, kobs, for the various sol-
vents studied in the Knoevenagel condensation of benzal-
dehyde and ethyl acetoacetate at 353 and 333 K catalysed by
IRMOF-3 are listed in Table 2.
Figure 1 also showed similar activities for both MeOH
(Table 2, entry 3) and EtOH (Table 2, entry 4) at 353 K,
which was notably above each of their boiling points under
atmospheric pressure. To achieve these observations the
reaction vessel was fitted with a reflux condenser while
being capped to conserve the static nitrogen atmosphere.
A second evaluationof MeOH (Table 2, entry 6) and EtOH
(Table 2, entry 7) as solvents was performed, this time at
333 K, which is lower than each of the solvents’ boiling
points. A similar trend to what was previously seen was
observed again at the lower temperatures, being that the more
The hydrophilic character of a catalyst support determines
the performance of supported homogeneous catalysts [5, 10].
A relationship occurs such that: the higher the hydrophilicity
of the catalyst support, the lower the effect of the solvent and
thus the faster diffusion can occur within the material. Not all
1
23

Knoevenagel Condensation Reactions
567
Table 2 Reaction conversions, calculated kobs and TOFs for the Knoevenagel condensation of benzaldehyde and ethyl acetoacetate catalysed by
IRMOF-
-
2
-1
-1
Initial TOF (min )
Run
Solvent
Temperature (K)
353
Reaction conversion (%)
k
obs (10
s
)
1
2
3
4
5
6
7
8
9
DMSO
DMF
64.5 ± 4.4
54.0 ± 0.5
43.0 ± 5.2
42.3 ± 1.4
15.8 ± 2.4
53.5 ± 2.2
34.7 ± 4.5
54.0 ± 0.0
39.0 ± 1.4
27.7 ± 4.3
27.6 ± 2.8
27.7 ± 1.9
10.53 ± 0.52
9.58 ± 0.43
7.38 ± 1.07
6.96 ± 0.92
1.17 ± 0.55
4.76 ± 0.50
2.32 ± 0.55
9.58 ± 0.43
2.11 ± 0.20
0.51 ± 0.38
0.52 ± 0.41
0.51 ± 0.38
0.44
0.39
0.22
0.29
0.15
0.39
0.29
0.39
0.27
0.10
0.10
0.08
MeOH
EtOH
Toluene
MeOH
EtOH
DMF
333
353
333
313
313
313
3
DMF
10
11
12
DMF
a
DMF
DMF
For all runs, the solvent volume was 5 cm and the values are the average of n = 3 repeats. Reaction conversions are reported at the end of the
run (130 min)
a
Benzaldehyde added first
polar solvent, MeOH, showed a higher activity with a con-
opposed to the addition of benzaldehyde first (Table 2, entry
11). This was performed to rule out the concept of the for-
mation of amides, where the ethyl acetoacetate would react
with the benzaldehyde, which would thus deactivate the
catalyst. No significant difference in the percent conversion
was observed in the catalytic ability of IRMOF-3. Gascon
et al. [5] similarly saw this occurring in the Knoevenagel
catalysed reaction of ethyl cyanoacetate and benzaldehyde.
A higher conversion might be expected when benzalde-
hyde is added first, as this would inhibit the formation of the
deactivating amides. However, this was not observed and the
conversion achieved by adding ethyl acetoacetate first was
version of 53.5 ± 2.2 % and k = (4.76 ± 0.50) 9
obs
-
2 -1
1
0
s . It was also noted that the conversion in MeOH was
raisedat333 K,incomparisontotheconversionat353 K.The
conversion in EtOH at 333 K, 34.7 ± 4.5 %, was lower than
its conversion at 353 K, 42.3 ± 1.4 %. However the rate of
-
2 -1
reaction was faster, kobs = (6.96 ± 0.92) 910
53 K than at 333 K where kobs = (2.32 ± 0.55) 910
Studies at 353, 333 and 313 K (Table 1, entries 8–10)
s , at
-2 -1
3
s .
were performed on the Knoevenagel condensation between
benzaldehyde and ethyl acetoacetate in DMF under a static
nitrogen atmosphere catalysed by IRMOF-3. DMF was
selected as the best solvent even though DMSO showed
higher activities at 353 K, but it was observed that the
amount of catalyst which could be recovered, by filtration
and washing with fresh DMF, after a catalytic run was
higher in DMF than it was in DMSO. Thus a sacrifice
between activity and recyclability was made in the aim to
optimise the catalytic reaction conditions of the catalyst.
Catalytic reactions performed at 353 K not only gave
higher conversions, 54.0 ± 0.0 % within the 130 min
where the reaction was monitored, but also a higher
-
2 -1
27.7 ± 1.9 % with kobs = (0.51 ± 0.38) 910
upon adding the benzaldehyde first the conversion attained
s , where
-
2 -1
s .
was 27.6 ± 2.8 % with kobs = (0.52 ± 0.41) 910
From the Arrhenius equation the activation energy, Ea,
can be calculated and states that:
ꢁEa
RT
k ¼ Ae
where k is the observed rate constant (s ), A the pre-
-
1
-
1
exponential factor, Ea the activation energy (J mol ),
-
1
-1
K ) and
R the Universal gas constant (8.3144621 J mol
-
2 -1
s
kobs = (9.58 ± 0.43) 910
was observed. At 333 K,
T the temperature (K). Thus a plot of 1/T versus ln kobs
produces a straight line where the product of the gradient
and the Universal gas constant equate to the Ea.
a lower temperature, a lower conversion of 39.0 ± 1.4 %
-
2 -1
was obtained with a kobs = (2.11 ± 0.20) 910
lowest temperature studied, 313 K, resulted in a conversion
s . The
From the Eyring-Polanyi equation the enthalpy, DH, and
entropy, DS, can be calculated and states that:
-
2 -1
s .
of 27.7 ± 4.3 % with a kobs = (0.51 ± 0.38) 910
To ascertain as to whether the order of addition was
important in the Knoevenagel reaction, between benzalde-
hyde and ethyl acetoacetate catalysed by IRMOF-3 in DMF
at 313 K, a comparative reaction was performed where the
ethyl acetoacetate (Table 2, entry 12) was added first as
ꢂ
kbT ꢁDG
RT
k ¼
e
h
-
where k is the observed rate constant (s ), k the Boltzmann
1
b
-
23
-1
J K ), T the temperature
constant (1.3806488 9 10
123

5
68
A. R. Burgoyne, R. Meijboom
Table 3 Reaction conversions, calculated kobs and TOFs in Knoevenagel condensation of benzaldehyde and ethyl acetoacetate or ethyl cya-
noacetate or diethylmalonate catalysed by IRMOF-3, UMCM-1-NH and MIXMOF
2
Run Substrate
Catalyst
Solvent
Temperature Reaction
(
k
K) conversion (%) (10
obs
Initial TOF
-1
(min )
Reference
-
2
-1
s
)
Ethyl acetoacetate
UMCM-1-NH
IRMOF-3
2
2
44.0 ± 3.4
27.7 ± 2.8
30.0 ± 2.1
73.1 ± 1.6
62.3 ± 1.7
62.9 ± 3.2
1.2 ± 0.4
99
3.04 ± 0.23
0.46 ± 0.40
0.55 ± 0.38
1.01
0.10
0.07
1
2
DMF
DMF
This work
This work
MIXMOF
Ethyl cyanoacetate
UMCM-1-NH
IRMOF-3
11.58 ± 0.13 2.02
2.87 ± 0.24
1.21
1.13
–
MIXMOF
3.26 ± 0.22
3
4
5
6
Diethyl malonate
Ethyl cyanoacetate
Ethyl acetoacetate
Ethyl cyanoacetate
IRMOF-3
DMF
–
–
–
–
This work
[5]
IRMOF-3
DMSO
DMSO
313
353
353
2.9
IRMOF-3
58
0.33
0.37
[5]
Ethyldiamine
grafted MIL-101
Cyclo
hexane
80
[17]
7
8
9
1
1
1
Ethyl acetoacetate
Ethyl acetoacetate
Ethyl acetoacetate
Ethyl acetoacetate
BEA
Xylene
Xylene
Xylene
Xylene
DMSO
403
403
403
403
413
413
82
41
40
58
52
53
–
–
–
–
–
–
–
–
–
–
–
–
[18]
[18]
[18]
[18]
[19]
[19]
TS-1
CuBTC
FeBTC
SBA-NH
0
1
2
0
2 -
2
-10-P
hydroxyacetophenone
C
6 5
H –
NO
2
1
3
4
C
6
H
5
–CN 413
35
12
–
–
–
–
[19]
[19]
1
Hexyl
alcohol
1,3,5-
CH
413
1
5
413
55
84
–
–
–
–
[19]
[19]
3
–
6
C H
3
1
6
–
413
Solvent DMF, n = 3 repeats
-
34
(
K), h the Planck’s constant (6.62606957 9 10
-
J s), DG*
the Gibbs energy of activation (J mol ), R the Universal gas
reaction to shed light on the catalytic abilities of basic
solids, particularly MOFs and zeolites [6].
1
-
1
-1
K
constant (8.3144621 J mol
K). Thus a plot of 1/T versus ln (k/T) produces a straight line
with gradient equal to –DH/R and with a y-intercept of ln
) and T the temperature
None of the MOF materials were able to catalyse the
Knoevenagel reaction between benzaldehyde and diethyl
(
malonate with any significant conversion, as the pK was
a
(
kb/h) ? DS/R.
The aforementioned plots were performed and E was
too high for the –NH groups present in the MOF. Selecting
2
a
a different substrate, ethyl acetoacetate, resulted in lowering
-
1
-1
found to be 973 J mol , DH was calculated as 933 J mol
the pK of the substrate from 13.3 to 10.7 allowed for sig-
a
-
1
-1
and DS as -84 J K mol for the IRMOF-3 catalysed
Knoevenagel condensation of benzaldehyde and ethyl ace-
toacetate in DMF.
nificantly higher results as shown in Table 3, which could
then be used for comparison of the different MOF catalysts.
For IRMOF-3 it was observed that only 1.2 ± 0.4 %
conversion was achieved when reacting diethyl malonate
with benzaldehyde; 27.7 ± 2.8 % conversion could be
achieved when reacting ethyl acetoacetate with benzalde-
hyde and 73.1 ± 1.6 % could be achieved when reacting
ethyl cyanoacetate with benzaldehyde after 130 min at
313 K in DMF. The difference in conversions is large, due
to the various strengths of the C–H bond of the methylenes
in the different substrates.
3
.2 Comparative Reactions Using IRMOF-3, UMCM-
-NH and MIXMOF
1
2
The ability of these basic MOFs to deprotonate different
methylene hydrogens was tested in the Knoevenagel cata-
lysed reaction between benzaldehyde and diethyl malonate
(
pK *13.3 [15]); ethyl acetoacetate (pK *10.7) and
For the same catalyst loading, as in IRMOF-3, a large
mass of MIXMOF was required. To provide 0.18 mmol –
a
a
ethyl cyanoacetate (pK *9.0 [16]) at 313 K in DMF. In
a
most cases, the Knoevenagel reaction is used as a test
NH in IRMOF-3 only 36 mg is required. However, for
2
1
23

Knoevenagel Condensation Reactions
569
Fig. 2 Knoevenagel
condensation of benzaldehyde
and ethyl cyanoacetate (dashed
dotted line) or ethyl
acetoacetate (dotted line) in
DMF at 313 K catalysed by
2
UMCM-1-NH (filled triangle);
MIXMOF (filled square);
IRMOF-3 (filled diamond);
n = 3 repeats
Table 4 Reaction conversions, calculated kobs and TOFs in Knoevenagel condensation of benzaldehyde and ethyl acetoacetate or ethyl cya-
noacetate catalysed by recycled IRMOF-3, UMCM-1-NH and MIXMOF
2
Run
Catalyst
Substrate
Solvent
Temperature (K)
313
Reaction conversion (%)
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
UMCM-1-NH
2
2
Ethyl acetoacetate
Ethyl cyanoacetate
Ethyl acetoacetate
Ethyl acetoacetate
Ethyl acetoacetate
Ethyl acetoacetate
DMF
DMF
DMF
DMF
DMF
DMF
44.0 ± 3.4
41.5 ± 2.7
39.9 ± 1.5
73.1 ± 1.6
72.6 ± 2.1
70.4 ± 1.9
27.7 ± 2.8
27.5 ± 1.8
25.8 ± 2.2
62.3 ± 1.7
60.9 ± 2.3
60.1 ± 2.5
30.0 ± 2.1
29.8 ± 1.8
27.6 ± 3.2
62.9 ± 3.2
61.7 ± 2.7
59.2 ± 1.9
UMCM-1-NH
IRMOF-3
IRMOF-3
MIXMOF
MIXMOF
313
313
313
313
313
Solvent DMF, n = 3 repeats
MIXMOF, as the percentage of ABDC is only 5 % amongst
the organic linkers, 720 mg MIXMOF is required to supply
ethyl acetoacetate, which only achieved 30.0 ± 2.1 % at
313 K in DMF after 130 min. MIXMOF and IRMOF-3 both
have similar catalytic abilities to catalyse the Knoevenagel
reaction between ethyl acetoacetate or ethyl cyanoacetate
and benzaldehyde.
0
.18 mmol –NH . This increase in mass made using
2
MIXMOF, as a catalyst, less desirable than IRMOF-3, as
twenty times the mass is required and the conversion of ethyl
cyanoacetate at 313 K in DMF was only 62.9 ± 3.2 % after
It is of note that UMCM-1-NH , with a larger surface
2
1
30 min. This was significantly larger than the conversion of
area and pore sizes, in both cases provided the highest
123

5
70
A. R. Burgoyne, R. Meijboom
-
2 -1
reaction rates with kobs = (11.58 ± 0.13) 910
the Knoevenagel condensation between ethyl cyanoacetate
s
for
for the recyclability study. The Knoevenagel condensation
of benzaldehyde and ethyl acetoacetate in DMF at 313 K
showed similar results, as shown in Table 4, for the first
three catalytic runs. GC analysis to quantify reaction con-
version for recyclability studies were performed only at
130 min for comparison of each catalytic run after recy-
cling of the catalyst. Upon completion of a catalytic run,
the catalyst was filtered from the reaction mixture and
washed in DMF at 313 K overnight and then washed with
CHCl . Activation of the catalyst and removal of CHCl
-
2 -1
s
and benzaldehyde and kobs = (3.04 ± 0.23) 910
for
ethyl acetoacetate and benzaldehyde. This large difference
in kobs, shown in Table 3, can be explained by the relative
ease by which the more acidic hydrogens of ethyl cya-
noacetate can be removed than the less acidic hydrogens of
ethyl acetoacetate. Of interest also are the similar yet
slightly larger kobs of MIXMOF, in both cases, compared to
that of IRMOF-3. This difference was analysed statistically
using the Student’s t test to compare the differences
between the data.
3
3
was achieved by drying the catalyst in a vacuum oven at
373 K at 560 mm Hg. Gascon et al. [5] also observed such
catalytic capabilities over the first three runs for IRMOF-3
in the Knoevenagel condensation of benzaldehyde and
ethyl cyanoacetate. Nguyen et al. [14] provided evidence
that for the first three catalytic runs ZIF-9 had the same
reaction rate in the Knoevenagel condensation of benzal-
dehyde and malononitrile.
UMCM-1-NH also requireda higher mass, thanIRMOF-3,
2
of catalyst to keep the catalyst loading constant. Although in
this case only 144 mg UMCM-1-NH is needed to provide
2
0
.18 mmol –NH . The catalytic ability of UMCM-1-NH2
2
was higher than IRMOF-3 and MIXMOF. UMCM-1-NH2
resulted in a larger conversion, over the 130 min, of
4
4.0 ± 3.4 % in the reaction between benzaldehyde and
ethyl acetoacetate and of 73.1 ± 1.6 % in the reaction
between benzaldehyde and ethyl cyanoacetate. These con-
versions are competitive with literature results, as shown in
Table 3. IRMOF-3 has been shown to be able to achieve a
remarkable 99 % conversion of substrate in DMSO at 313 K
when ethyl cyanoacetate was used as a substrate and with
ethyl acetoacetate only 58 % was observed [5]. Ethyldi-
amine grafted MIL-101 yielded 80 % reaction conversion of
ethyl cyanoacetate at 353 K [17], which is higher than all the
observed reaction conversions in this study although the
solvent DMSO was used at 353 K, which was higher than the
studies performed on ethyl cyanoacetate. In the work of
Opanasenko et al. [18] BEA, TS-1, CuBTC and FeBTC were
shown to give 82, 41, 40 and 58 %, respectively, in the cat-
alysed reaction of benzaldehyde and ethyl acetoacetate at
4 Conclusions
In this study IRMOF-3, UMCM-1-NH2 and MIXMOF
were compared as basic heterogeneous catalysts in the
Knoevenagel condensation between benzaldehyde and
ethyl acetoacetate or ethyl cyanoacetate. IRMOF-3 was
found to yield higher conversion rates than the homoge-
neous analogue, aniline. This suggests the idea that the
basicity of the amine group within the MOF structure is
enhanced when within the MOF material [5].
UMCM-1-NH in all cases had higher catalytic activities
2
than IRMOF-3 and MIXMOF which had near identical
catalytic activities in all cases. Optimization of the reaction
conditions found that 353 K was the optimal temperature for
higher substrate conversion, but lower temperatures, 313 K,
allowed for better catalyst recovery. Substrates with lower
pKa values yielded high conversions. From the Arrhenius
4
03 K in xylene. Although, in this study IRMOF-3 and
MIXMOF did result in lower reaction conversions than the
study performed by Opanasenko et al., this is most likely due
to the temperature it was performed at, 313 K, as opposed to
-
1
equation it was found that E = 973 J mol and from the
a
-
1
4
03 K. Similarly, the amine containing SBA-15 zeolite was
Eyring-Polanyi equation that DH = 933 J mol
and
-
1
-1
shown to have basic catalytic abilities in catalysing the car-
0
DS = -84 J K mol for the IRMOF-3 catalysed Knoe-
venagel condensation between benzaldehyde and ethyl
acetoacetate.
bon coupling of the 2 -hydroxyacetophenone and benzalde-
hyde, as shown in Table 3, in a variety of solvents. The
reaction conversions were of a similar magnitude to the
results obtained in this study, despite the lower surface
area of the zeolitic basic catalyst and the higher reaction
temperature, 413 K [19] (Fig. 2).
Acknowledgments Funding from the South African NRF, Sasol
R&D and the research fund from the University of Johannesburg are
gratefully acknowledged. ARB is grateful to the NRF-DAAD for a
scholarship. Mr. D. Harris and Dr. R. Meyer (Shimadzu South Africa)
are acknowledged for use of their equipment.
3
.3 Recyclability
References
A large loss of mass, up to as much as half the mass in
some cases due to attrition and dissolution, of the hetero-
geneous catalyst at 353 K and DMSO as a solvent was
observed. Consequently, less harsh conditions were chosen
1
. Derouane EG, Roberts SM (2007) Catalysts for fine chemical
synthesis, microporous and mesoporous solid catalysts. Wiley,
West Sussex
1
23

Knoevenagel Condensation Reactions
571
2
3
4
5
6
7
8
9
. Ernst S, Hartmann M, Sauerbeck S, Bongers T (2000) Appl Catal
A 200:117–123
. Rao YVS, De Vos DE, Jacobs PA (1997) Angew Chem Int Ed
12. Corma A, Iborra S, Rodriguez I, Sanchez F (2002) J Catal
211:208–215
13. Climent MJ, Corma A, Dominguez I, Iborra S, Sabater MJ, Sastre
G (2007) J Catal 246:136–146
14. Nguyen LTL, Le KKA, Truong HX, Phan NTS (2012) Catal Sci
Technol 2:521–528
15. Brown WH, Foote CS, Iverson BL, Anslyn EV, Novak BM
(2011) Organic chemistry, 6th edn. Brooks/Cole, Belmont
16. Sheldon RA, van Bekkum H (2008) Fine chemicals through
heterogenous catalysis. Wiley, Weinheim
17. Hwang YK, Hong DY, Chang JS, Jhung SH, Seo YK, Kim J,
Vimont A, Daturi M, Serre C, Ferey G (2008) Angew Chem Int
Ed 47:4144–4148
18. Opanasenko M, Dhakshinamoorthy A, Shamzhy M, Nachtigall P,
ˇ
Hor a´ cˇ ek M, Garcia H, Cejka J (2013) Catal Sci Technol
3
6:2661–2663
. Solomons TWG, Fryhle C (2009) Organic chemistry, 10th edn.
Wiley, New York
. Gascon J, Aktay U, Hernandez-Alonso MD, van Klink GPM,
Kapteijn F (2009) J Catal 261:75–87
. Corma A, Fornes V, Martin-Aranda RM, Garcia H, Primo J
(
1990) J Appl Catal 59:237
. Kleist W, Maciejewski M, Baiker A (2010) Thermochim Acta
99:71–78
. Koh K, Wong-Foy AG, Matzger AJ (2008) Angew Chem Int Ed
7:677–680
. Freeman F (1980) Chem Rev 80:329–350
4
4
1
1
0. Tietze LF (1996) Chem Rev 96:115–136
1. Macquarrie DJ, Jackson DB, Clark JH (2001) In: Sherrington DC,
Kybett AP (eds) Supported catalysts and their application. RSC,
Cambridge
3:500–507
19. Mart ´ı n-Aranda RM, Cejka J (2010) Top Catal 53:141–153
ˇ
123