A designed organic-zeolite hybrid acid-base catalyst

Article abstract of DOI:10.1016/j.jcat.2011.09.004

An organic-zeolite hybrid catalyst was synthesised via solid-state impregnation of the zeolite with an amount of melamine corresponding to 20 mol% of the aluminium content. A high density of basic sites is formed in the zeolite. From infrared spectroscopy and TGA measurements we infer that melaminium cations are formed in the zeolite which are highly thermally stable. IR spectroscopic and TGA measurements showed the presence of melaminium after pre-treatment in vacuum at 450 °C. The catalyst exhibited activity in the Knoevenagel condensation reaction between benzaldehyde and ethyl cyanoacetate which is suggested to be promoted by the presence of acid and base sites. The catalyst activity was compared with other known base catalysts including hydrotalcites and magnesium oxide on alumina support.

Full text of DOI:10.1016/j.jcat.2011.09.004

Journal of Catalysis 285 (2012) 10–18
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
A designed organic–zeolite hybrid acid–base catalyst
Andrew C. Brooks a, Liam France b, Cecile Gayot b, Jerry Pui Ho Li b,c, Ryan Sault c, Andrew Stafford c,
John D. Wallis a, Michael Stockenhuber b,c,
⇑
a Natural Science Research Centre, School of Science and Technology, Nottingham Trent University, Clifton Lane, Nottingham NG11 8NS, UK
b The Catalysis and Nanoscience Laboratory, School of Science and Technology, Nottingham Trent University, Clifton Lane, Nottingham NG11 8NS, UK
c Catalysis Research Laboratory, Priority Research Centre for Energy, Chemical Engineering, School of Engineering, The University of Newcastle, Callaghan Campus, NSW 2308, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
An organic–zeolite hybrid catalyst was synthesised via solid-state impregnation of the zeolite with an
amount of melamine corresponding to 20 mol% of the aluminium content. A high density of basic sites
is formed in the zeolite. From infrared spectroscopy and TGA measurements we infer that melaminium
cations are formed in the zeolite which are highly thermally stable. IR spectroscopic and TGA measure-
ments showed the presence of melaminium after pre-treatment in vacuum at 450 °C. The catalyst exhib-
ited activity in the Knoevenagel condensation reaction between benzaldehyde and ethyl cyanoacetate
which is suggested to be promoted by the presence of acid and base sites. The catalyst activity was com-
pared with other known base catalysts including hydrotalcites and magnesium oxide on alumina
support.
Received 25 June 2011
Revised 31 August 2011
Accepted 4 September 2011
Available online 26 October 2011
Keywords:
Base zeolite
Organic–inorganic hybrid
In situ IR spectroscopy
Knoevenagel condensation
High thermal stability
Crown Copyright Ó 2011 Published by Elsevier Inc. All rights reserved.
1. Introduction
introduced and stabilised in zeolites. They showed significant basi-
city [4,5] but exhibited limited stability, site accessibility, diffusion
Base catalysed reactions are an important fundamental reaction
type in modern organic synthesis. Classic base catalysts are NaOH,
KOH, alkali metal (mainly sodium) alcoholates or LDA (lithium
diisopropylamide). The catalysts mentioned are all used as homo-
geneous catalysts, with the usual difficulties of separation, waste
disposal and regeneration. Compared to the broad application of
acidic zeolites as solid catalysts in chemical processing technology,
there has been much less attention paid to basic microporous and
mesoporous materials, despite the considerable potential for use in
a number of industrially important reactions including aldol addi-
tions, condensations and transesterification reactions [1]. Out of all
industrial applications of solid acid and base catalysts, only 8% of
the processes utilise solid base catalysts, with none of them being
basic zeolite catalysts [2]. It is possible to create basic zeolite cat-
alysts by impregnation and ion exchange of alkali metal salts.
The basicity of these materials can be tailored by variation of coun-
terion [3] and also aluminium content. These zeolites, however,
possess base sites of comparably low strength which limits their
applicability to organic synthesis, though the base sites are ex-
pected to be more resistant to poisoning from water or carbon
dioxide. Alkali metals, metal azides and oxides have also been
limitation and coking.
There is thus an incentive to develop active and selective heter-
ogeneous catalysts for a number of base-catalysed reactions. De-
spite some success in the development of heterogeneous base
catalysts, such as hydrotalcites [6], MgO [7], alkaline-exchanged
zeolites [8], alumina [9] and mesoporous materials [10], the variety
and success are still limited. Hydrotalcites, MgO, zeolites and basic
alumina exhibit basicity which is related to the bonding and
charges on the framework and extra framework oxygen atoms. Or-
ganic–inorganic hybrid catalysts have also been reported [11] and
though are found to be less strongly basic than the corresponding
free organic molecules. Base catalysts can also be created by the
introduction of nitrogen-containing groups [12], by first covalently
bonding trimethoxysilyl-propyl-N,N,N-trimethylammonium chlo-
ride (SiNR4Cl) to the silanol groups of the solid catalyst, before
reacting with the organic bases of interest. Using a diamino-func-
tionalised mesoporous silica, Choudary et al. [13] showed how a
Knoevenagel condensation between 3,4,5-trimethoxybenzalde-
hyde and malononitrile took place on a base-modified surface.
Due to its open structure, zeolite beta was shown to exhibit excel-
lent rates and diffusion properties for a number of acid-catalysed
reactions. Its use as a base catalyst, however, was only recently
shown. Some success in producing a basic zeolite beta catalysts
was reported by Ernst et al. who used a nitridation method [14].
It was found that such catalysts are active for the Knoevenagel con-
densation reaction between benzaldehyde and malononitrile and
⇑
Corresponding author at: Catalysis Research Laboratory, Priority Research
Centre for Energy, Chemical Engineering, School of Engineering, The University of
Newcastle, Callaghan Campus, NSW 2308, Australia.
E-mail address: michael.stockenhuber@newcastle.edu.au (M. Stockenhuber).
0021-9517/$ - see front matter Crown Copyright Ó 2011 Published by Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.09.004

A.C. Brooks et al. / Journal of Catalysis 285 (2012) 10–18
11
that there is a correlation between the degree of benzaldehyde
conversion and the amount of nitridation of the zeolite beta cata-
lyst [14]. To overcome diffusional limitations observed in zeolites,
mesoporous materials have been tested in condensation reactions
with varying success. Recently, zeolites with high mesoporosity
have been synthesised and tested for condensation reactions
[15,16].
Corma et al. [17] suggest that by changing the chemical compo-
sition and activation conditions of the base catalyst, it is possible to
have predominantly Lewis or Brønsted base catalysts within a
large range of well-defined basicities. Catalysts that contain mild
acid–base pairs, such as those that exist in amorphous alumino-
phosphates (ALPO), have been shown to achieve high selectivity
in Knoevenagel condensation of heptanal and benzaldehyde.
The Knoevenagel condensation reaction between ethyl cya-
noacetate and benzaldehyde has long been recognised as a rela-
tively simple test for the effectiveness of heterogeneous basic
catalysis. Ethyl cyanoacetate has two highly acidic protons in the
boat and heated under a nitrogen atmosphere to 300 °C. The sam-
ple was kept at 300 °C for 24 h.
2.1.3. Hydrotalcite catalyst
Hydrotalcite with an Mg/Al ratio of 2 was prepared via co-pre-
cipitation [6]. Two aqueous solutions were prepared, one 70 ml
containing 0.35 mol NaOH and 0.09 mol Na2CO3 (Aldrich reagent-
plus P98%), and the other 45 ml of 0.1 mol Mg(NO3)2ꢀ6H2O (Al-
drich P98%) and 0.05 mol Al(NO3)3ꢀ9H2O (Aldrich P98%). These
solutions were slowly added together over one hour under vigor-
ous stirring. The resulting white suspension was heated to 60 °C
and continuously stirred for 24 h. The precipitate was then sepa-
rated using a vacuum filter and washed extensively with distilled
water. The sample was then dried at 132 °C for 20 h.
The hydrotalcite catalyst was activated by placing the material
in a controlled atmosphere furnace and heated under nitrogen at
5 °C/min to 450 °C for a period of 8 h. It was then left to cool to
room temperature. The hydrotalcite was subsequently rehydrated
by suspending it in decarbonised water under nitrogen atmosphere
for 1 h. The sample was filtered under nitrogen atmosphere and
washed thoroughly with decarbonised water and dried in nitrogen
flow at 25 °C.
a
position to the nitrile (pKa = 9) [18]. Thus, catalysts of relatively
low base strength can be used for the reaction, such as lithium, so-
dium, potassium and caesium X and Y type zeolites [19].
In this communication, we report for the first time the modifi-
cation of a microporous zeolite beta with melamine to form an
inorganic–organic hybrid catalyst. Melamine is an organic mole-
cule of medium basicity (pKa = 5) but also possesses high thermal
stability [20,21]. This allows for higher activation temperatures.
Of interest is the mechanism by which the conversion takes place.
We have deliberately chosen the open structure of zeolite beta,
whose open pores have a sufficient pore diameter (average meso-
pore distribution of 1.5–2.0 nm) [22–24] for the best chance of
deposition of our base which has a kinetic diameter of 0.62 nm
[25]. The presence of micropores allows for good shape selectivity,
while the open pore structure is allowing for good interaction with
reactants. Our synthesis method for the catalyst is also compara-
tively simple, compared with other modification methods, as it re-
quires considerably less time and milder conditions. The addition
of melamine to the zeolite provides additional base sites to pro-
duce acid–base pairs. We also compare the catalytic activity ob-
served in the Knoevenagel condensation of benzaldehyde with
ethyl cyanoacetate using this newly developed material, with
two other known solid base catalysts, hydrotalcite and MgO/Al2O3.
2.1.4. MgO/Al2O3 catalyst
MgO/Al2O3 was prepared by grinding MgO (Aldrich P98%) and
Al2O3 (Aldrich P99%) together in a 1:1 M ratio using a mortar and
pestle. A small amount of distilled water was added and mixed to
form a thick slurry. The sample was then dried at 132 °C for 12 h.
After drying, the sample was crushed into a powder and heated
in a controlled nitrogen atmosphere furnace to 200 °C and held
there for 8 h. The heating rate from room temperature to 200 °C
was 5 °C/min.
2.2. Catalytic experimental setup
Prior to reaction, the sample was activated under a nitrogen
atmosphere at 450 °C for 1 h, cooled to 77 °C and transferred into
a 100 mL glass reactor, which was purged with dry nitrogen. To
minimise water adsorption, the catalyst was transferred hot into
the vessel. The reaction was performed in the liquid phase at
80 °C under a nitrogen atmosphere, under solventless conditions
using an excess of benzaldehyde. The 100-mL glass reactor con-
tained 0.2 mol benzaldehyde (Aldrich reagentplus P99%),
0.117 mol ethyl cyanoacetate (Aldrich P98%), 0.054 mol toluene
(Aldrich reagentplus 99%) and typically 300 mg of catalyst. The tol-
uene was added as an internal standard.
2. Experimental
2.1. Catalyst preparation
2.1.1. Melamine zeolite catalyst, ‘‘mel-20-beta and mel-100-beta’’
The melamine zeolite catalyst was prepared by thoroughly
grinding melamine (99% SigmaAldrich) with an ammonium ex-
changed zeolite beta (Zeolyst, Si/Al = 12.5, NH4-beta) in a ball mill
for 24 h. The amount of melamine added to the zeolite beta is
equivalent to 20 mol% or 100 mol% of the aluminium present in
zeolite beta. Catalysts were named mel-20-beta and mel-100-beta,
respectively. The rationale to prepare differently loaded catalysts
was to produce a purely basic catalyst (mel-100-beta) and a cata-
lyst that contains acid and base sites (mel-20-beta). The mixture
was transferred into a quartz calcination boat and heated under
a nitrogen atmosphere to 300 °C. The sample was kept at 300 °C
for 24 h.
The reaction components were quantified using gas chromatog-
raphy (GC). Two GCs were used for the analysis. The first one was a
Shimadzu GC8A equipped with a 3-m-long stainless-steel, 10% OV-
17 liquid-phase WHP-supported column. Separation was carried
out using a temperature program with the range set at 80–
350 °C. The reaction samples were analysed using gas chromatog-
raphy (GC). The second instrument used for the analysis was a Shi-
madzu GC-2010 equipped with a 30-m fused silica RtxÒ-5 column
with a 5% diphenyl/95% dimethyl polysiloxane stationary phase
equipped with an FID detector.
Small samples were taken from the reaction vessel at regular
intervals over a period of 24 h. Samples were then injected into
the GC for analysis. Separation was carried out using a temperature
program with the range set at 353–623 K using a ramp rate of 16 K/
min, with the carrier gas set to 2 ml/min, on a 100:1 split setting.
The final temperature was held for 10 min. Toluene was used as
an internal standard, and product and reactants were calibrated
using standards. The error on the conversions and yields was found
to be 1% absolute.
2.1.2. Acid catalyst
Zeolite H-beta was used as an acid catalyst. This form of the cat-
alyst was made by calcining ammonium exchanged zeolite beta
(Zeolyst, Si/Al = 12.5, h-beta) at 480 °C for 5 h. Upon reaction, the
catalyst was activated by transferring it into a quartz calcination

12
A.C. Brooks et al. / Journal of Catalysis 285 (2012) 10–18
O
353 K
O
CN
H
NC
O
H2O
+
+
Zeolite
OEt
catalyst
H
OEt
Scheme 1. The Knoevenagel condensation between benzaldehyde and ethyl cyanoacetate.
Scheme 2. The Michael addition of ethyl cyanoacetate to ethyl trans-a-cyanocinnamate.
2.3. Characterisation
envisaged as indicated in Scheme 3. There is overwhelming evi-
dence from the Cambridge Structural Database [30] that melamine
is preferentially protonated on a ring nitrogen atom as shown in
over 65 structures of its monocation salts such as its chloride, sul-
phate and dimesylamide salts [31–33]. Indeed, there are cases
where a dication is formed on protonation and both the protons
are located on ring nitrogen atoms, as in its bis(trifluoroacetate),
sulphate and bis(perchlorate salts) [34–36].
Gravimetric analysis was conducted using a thermal gravimet-
ric analyser (TGA) from SetaramÒ, model Setsysev™ 1200. The zeo-
lite sample was heated inside the TGA chamber under an inert
nitrogen atmosphere. Mass changes were observed as a function
of temperature. The surface area and micropore volume of some
catalysts were measured by nitrogen adsorption at 77 K using
Gemini 11 2370 surface area analyser. Fourier transform infrared
spectroscopy (FTIR) was undertaken using an ATI Research Series
1 FTIR Spectrometer. The catalyst was compressed into a disc using
a pellet press, activated at 450 °C and then left to cool to 50 °C. De-
tails of the in situ experimental setup can be found elsewhere [26].
3.1. Characterisation
3.1.1. TGA analysis
Fig. 1 shows TGA analysis of the NH4-beta zeolite together with
the differential TGA (DTGA). The findings are summarised in Ta-
ble 1. Differential TGA analysis shows that at least two distinct
steps can be observed, one centred at 130 °C with a corresponding
4.5% weight loss and one broad maximum at 390 °C with a corre-
sponding weight loss of 2.65% up to 510 °C. The peak at 130 °C
can be attributed to desorption of weakly adsorbed water [37],
whereas the temperature range of the second weight loss 300–
500 °C suggests removal of ammonia from the ammonium form
of the zeolite. The weight loss is slightly higher than what would
be expected under the assumption of a 1:1 ratio between tetrahe-
dral aluminium and the ammonium cation (1.23 mmol gꢁ1).
Fig. 2 shows the TGA and DTGA curves for the melamine ex-
changed zeolite (mel-20-beta). Again peaks at lower and higher
temperatures were observed in the differential TGA. The low tem-
perature peak (89 °C) is at a slightly lower temperature than the
corresponding peak of the NH4-beta zeolite but the position of
the high temperature weight loss (395 °C) is similar. The steps in
the TGA curves are significantly more distinct compared with the
NH4-beta zeolite. The amount of weight lost at higher tempera-
tures, attributed to decomposition of ammonium to ammonia, is
3. Results and discussion
Scheme 1 shows the reaction equation for the Knoevenagel con-
densation of benzaldehyde with ethyl cyanoacetate to form ethyl
trans-a-cyanocinnamate. The reaction often is carried out in DMSO
as a solvent, mainly to reduce the undesired consecutive Michael
addition side reaction illustrated in Scheme 2. However, over our
selective solid-state catalyst (mel-20-beta), the reaction can be
performed under solventless conditions without the complication
of solvent recovery and no formation of Michael by-product. For
cation-exchanged zeolites, the reactivity for the Knoevenagel con-
densation reaction tends to increase with cation size [27] and alu-
minium content of the zeolite. Generally, these two parameters are
associated with basic strength of the catalyst [16,28]. Catalytic
activity was shown to exhibit a complicated pattern with activity
maxima dependent on basic site content, strength and crystallinity
of the framework and accessibility of the sites in a number of basic
catalysts [29]. For example, Cs exchanged SBA-15 which is a highly
basic material showed less activity for the condensation reaction
than framework nitride zeotypes which are expected to exhibit
lower basic strength. Thus, it appears that intermediate basicity
is beneficial for the activity of the catalyst. Furthermore, the con-
densation reaction occurs over sites with moderate basicity, and
lower basicity often decreases deactivation. We have thus chosen
a relatively weak base (melamine pKa = 5) as our modifier for the
zeolite. Melamine can react with the zeolitic acid sites, thereby di-
rectly anchoring it to the microporous zeolite beta pore surface by
electrostatic interaction and hydrogen bonding. The other nitrogen
atoms are then able to form the basic site for the catalytic reaction.
The possible bonding between the melamine and the zeolite can be
Scheme 3. Structure of the active site.

A.C. Brooks et al. / Journal of Catalysis 285 (2012) 10–18
13
Fig. 1. TGA plot of sample mass (mg) vs temperature (°C), for the NH4-beta zeolite catalyst.
very similar for mel-20-beta and NH4-beta zeolites. The similar
3.1.2. In situ FTIR spectroscopy
weight loss suggests that the ammonium ions are still present in
the modified catalyst before heat treatment. The high temperature
weight loss we observe for both catalysts is higher than would be
expected for the aluminium content of the zeolite. Similar behav-
iour has been found for other zeolites [38]. The main difference be-
tween mel-20-beta zeolite catalyst and NH4-beta zeolite is a less
sharp decomposition peak for the NH4-beta zeolite compared with
mel-20-beta zeolite. Although the desorption is occurring over a
wider temperature range and is less distinct, the overall weight
loss is the same for the two different materials. Broadening of
the peak can be explained by a broader energy distribution of the
desorption enthalpy. It is expected that the melamine is adsorbing
on the stronger sites, consistent with a reduced weight loss at a
higher temperature. We can also assume that the ammonium ions
are still present after grinding the catalyst with the melamine. It is
thus likely that the NHþ4 cation forms an adduct with the bound
melamine. The temperature treatment is suggested to desorb the
NH3 and result in protonation of the melamine to form the cation
observed by IR spectroscopy. As evidenced by TGA analysis, ther-
mal decomposition of melamine or melaminium does not take
place when the melamine is anchored in the zeolite (since the
weight loss we observe is too small). Decomposition of melamine
is reported to begin at 500 °C [20]. The absence of significant
weight changes at temperatures higher 500 °C suggests stabilisa-
tion of the melamine in the zeolite. The weight loss at 395 °C
and infrared spectroscopy (see below) are consistent with the pres-
ence of a melaminium cation which is expected to exhibit signifi-
cantly higher stability than the melamine.
Fig. 3 shows the in situ FTIR spectra of the parent H-beta zeolite
and the mel-20-beta catalyst. For the parent H-beta zeolite, bands
were seen at 3745, 3609, 1992, 1875, 1713 and 1675 cmꢁ1. The
band at 3745 and 3609 cmꢁ1 corresponds to stretching vibrations
of terminal silanols and bridging hydroxyl groups, respectively.
The 1986 and 1872 cmꢁ1 bands correspond to overtones of the
Si–O stretching bond vibrations [39].
Bands at 3742, 3606, 3378, 3241, 1994, 1870, 1555, 1479 and
1425 cmꢁ1 were observed for the activated mel-20-beta sample
as well as bands with a central peak at approximately 2240 cmꢁ1
(shoulders at 2306, 2239 and 2279 cmꢁ1). The bands in the 3742
and 3606 cmꢁ1 region originate from the hydroxyl species of the
zeolite framework, associated with terminal O–H groups, and
bridging hydroxyls, respectively. This is supported by other FTIR
studies reported in literature [14,40,41]. It should be pointed out
also that the band at 3606 cmꢁ1 when compared with the parent
zeolite (the H-beta zeolite) appears to have undergone a red shift,
indicating that some interaction has occurred with the bridging
hydroxyls present in the zeolite. We observed bands between
1994 cmꢁ1 and 1870 cmꢁ1 that are also observed in the parent zeo-
lite. These bands are attributed to the overtone bands of the Si–O
stretching vibrations reported previously [39]. The bands have
been used to normalise for wafer thickness as outlined in [42].
Bands in the 3378 and 3241 cmꢁ1 region are attributed to NH
stretching vibrations of the side chain primary amines of the mel-
amine, similar to other melamine-derived compounds [43]. The
existence of these vibrations shows that the side chain amines in
the melamine molecule have not bonded with the zeolite surface.
Observed bands at 1655 cmꢁ1, 1557 cmꢁ1 and 1520 cmꢁ1 asso-
ciated with the ring structure of the melamine species were also
observed on the mel-20-beta catalyst. 1635 cmꢁ1, 1550 cmꢁ1 and
1520 cmꢁ1 bands are attributed to the C–N stretching vibrations
of the ring nitrogens. The band at 1635 cmꢁ1 is attributed to the
stretching modes of the aromatic C–N bonds from the ring. Bands
in the range of 2240 cmꢁ1 can be attributed to vibrations of the
amine cation, which are N–H stretching and overtones or combina-
tion bands in Fermi resonance. The position of the band at lower
wavenumbers suggests that the melamine binds as an amine salt
[44]. The IR absorptions are quite unique because only a few
Table 1
Adsorption properties of zeolite NH4-beta and mel-20-beta. Please see text for
discussion of ammonia desorption.
Zeolite
catalyst
NH3 loss at
temperatures
>300 °C
Micropore
volume
Langmuir
External
surface area
surface area
(cm3 gꢁ1
)
(m2 gꢁ1
)
(m2 gꢁ1
)
(mmogꢁ1
)
NH4-beta
mel-20-beta
1.86
1.90
0.160
0.127
731
606
185
163

14
A.C. Brooks et al. / Journal of Catalysis 285 (2012) 10–18
Fig. 2. TGA plot of sample mass (mg) vs temperature (°C), for the mel-20-beta zeolite catalyst.
Fig. 3. FTIR spectra comparing H-beta and mel-20-beta zeolite catalysts. Both samples were activated in vacuum (<10ꢁ7 mbar) at 450 °C.
compounds absorb IR in this region. The NH+ stretching frequency
is a little lower than was observed for halogen salts of amines but
is known to depend on the type of counterion. The frequency was
found to increase progressively from Clꢁ to Brꢁ and to Iꢁ and was
suggested to be associated with hydrogen-bonding interaction
with the counteranion [44,45]. Estimation of the partial charge
on the cation depending on the type of anion (Iꢁ, Brꢁ, Clꢁ and Iꢁ)
using a Sanderson electronegativity model suggests an increase
in the cation partial charge on the nitrogen in the amine and cor-
responding decrease in the frequency. The partial charge on a
NHþ4 cation bonded to the beta zeolite is estimated at 0.17, whereas
for NH4I, it is 0.155. This suggests a decrease in the NH+ frequency
in the melaminium cation due to the zeolite ‘‘anion’’, which is what
we observe by comparison with the literature. A paper on the de-
tails of the interaction between melamine and the zeolite is in
preparation.
Bands at 1479 cmꢁ1 and 1425 cmꢁ1 are associated with the pos-
itively charged amine deformation vibrations from the melamine.
The bands show evidence of N–H bending mode and two N–H
stretching modes, respectively [46], indicating that a bond be-
tween the zeolite and the amine cation from melamine has been
formed.

A.C. Brooks et al. / Journal of Catalysis 285 (2012) 10–18
15
100
80
60
40
20
0
4
3
2
1
0
0
100
200
300
0
100
200
300
400
Time / min
Time / min
Fig. 5. Percentage yield product (ethyl trans-a-cyanocinnamate) using the acid
catalyst H-beta zeolite as a function of time.
Fig. 4. Percentage yield (ethyl trans-a-cyanocinnamate) as a function of time over
the mel-20-beta zeolite catalyst.
4
3
2
1
0
Table 2
Product % conversion over 24 h of ethyl cyanoacetate and the respective rates where
applicable.
Zeolite catalyst Product conversion after 24 h (%) Initial rate (mol/g/min)
mel-20-beta
mel-100-beta
H-beta
93.0
18.0
13.8
3.5a
0.0015
0.0005
0.0008
–
Uncatalysed
0
100
200
300
400
a
Time / min
Michael addition product.
Fig. 6. Percentage yield product (ethyl trans-a-cyanocinnamate) without catalyst
as a function of time.
The preparation procedure involves grinding of melamine with
the NH4-beta zeolite and then heating to 450 °C in the infrared
spectrometer. It has been suggested that melamine begins to
decompose thermally at 500 °C and can form extended condensa-
tion products [20]. The formation of these products is expected to
be hindered in the zeolite pores. We also observe the characteristic
melaminium cation bands after pretreatment at 450 °C, resulting
in stabilisation through formation of melaminium cations.
Our results obtained for the mel-20-beta zeolite are comparable
to a report by Zhang et al. [47] whose study involved the grafting of
amino groups onto zeolite X variants. They observed a 55% conver-
sion over 50 min, indicating that the amino groups were key to cat-
alysing the reaction. Our values compare very well with a recently
reported system, layered PLS-1 [48], which exhibited similar end
conversions at 70 °C. After 30 min, 80% conversion was observed
in toluene. The initial rate based on the data given was
1.3 ꢂ 10ꢁ3 mol minꢁ1 gꢁ1. As can be observed from the above data,
over the mel-20-beta zeolite catalyst, no other reactions are ob-
served and the rate of reaction is significantly enhanced compared
with the other materials. Scheme 4 shows the mechanism for the
acid catalysed condensation reaction. In Scheme 5, the base-cata-
lysed Knoevenagel condensation is shown, which is probably oper-
ating over the mel-100-beta catalyst. The acid-catalysed reaction is
suggested to occur via protonation of the benzaldehyde by the zeo-
litic protons and attack on the resulting cation by the enol form of
the cyanoester. This nucleophilic reaction yields an alcohol, which
3.2. Catalytic activity of zeolite catalysts
Fig. 4 shows the percentage yield of ethyl trans-a-cyanocinna-
mate based on ethyl cyanoacetate against time for the mel-20-beta
zeolite catalyst. No by-product was observed. The initial rates and
final conversions over the various catalysts are given in Table 2. It
is important to note that we observed significantly lower conver-
sion and rate over mel-100-beta than over mel-20-beta. This sug-
gests changes in the site type and/or mechanism as outlined
below. A change in the micropore volume upon modification was
observed (Table 1); thus, changes in accessibility variation of the
sites in mel-20-beta and mel-100 beta cannot be ruled out. The
can be dehydrated to form the ethyl trans-a-cyanocinnamate.
Since the mel-20-beta catalyst has both protons and melaminium
sites present, it is suggested that the reaction occurs via a com-
bined mechanism as shown in scheme 6.
final conversion at 24 h of ethyl trans-a-cyanocinnamate over
mel-20-beta was observed to be 93%. The reaction observed using
H-beta zeolite catalyst produced a conversion of 13.8% (Fig. 5),
with minimal Michael addition product observed, due to the steric
hindrance induced from the zeolite pores. The acidic nature of the
catalyst, however, makes the formation of the desired product dif-
ficult. Similar, we observed 18% conversion over the more basic
zeolite mel-100-bea. Thus, the intermediate basicity of zeolite
mel-20-beta appears to favour the reaction to the desired product.
A conversion of 3.5% after 5 h (Fig. 6) was observed for the uncatal-
ysed reaction with significant formation of the Michael product.
The acid-catalysed reaction was found to exhibit a significantly
lower initial rate than our base-catalysed reaction, namely 0.0008
and 0.0015 mol gꢁ1 minꢁ1, respectively. Final conversion after 22 h
was 11.7% and some benzoic acid formation has been observed as
3.3. Catalyst comparisons
We observed that hydrotalcite effectively catalysed the Knoeve-
nagel condensation of ethyl cyanoacetate and benzaldehyde with
the conversion of 79.7% at 300 min. The results are shown in
Fig. 7 for the first 300 min. It can be seen that these values are low-
er than those reported in the literature [25] which showed conver-
sions of 31% and 88% after 15 min and 240 min, respectively. These
values were found with an Mg/Al ratio of three at 60 °C with
5.5 wt% catalyst. As an increase in Al content results in a higher
density of basic sites, it would be expected that the Mg/Al ratio
of 2:1 and 80 °C would achieve higher conversions than the Mg/
Al ratio of 3:1 at 60 °C. The layered structure of the hydrotalcite
well. Some ethyl trans-a-cyanocinnamate as well as some benzoic
acid was formed even without the presence of a catalyst.

16
A.C. Brooks et al. / Journal of Catalysis 285 (2012) 10–18
Scheme 4. Acid-catalysed condensation reaction (with H-beta zeolite catalyst).
Scheme 5. Base-catalysed condensation reaction with mel-100-beta catalyst (above) and using the zeolite oxygen anion as the basic catalyst (below).
can explain the selectivity towards product formation, as no Mi-
chael addition product was seen.
hyde. Conversions of 63.2% and 99.8% were obtained after 300
and 1425 min, respectively.
Fig. 8 shows the result obtained for the MgO/Al2O3 catalyst. We
observed product formation from using MgO/Al2O3 catalyst for the
Knoevenagel condensation of ethyl cyanoacetate and benzalde-
The modification of MgO with Al2O3 creates a more basic cata-
lyst, which would be expected to perform better than MgO due to
an increased surface area. This idea is supported by a recent report

A.C. Brooks et al. / Journal of Catalysis 285 (2012) 10–18
17
Scheme 6. Proposed mechanism for the synergistic relationship between the acidic and basic sites of the mel-20-beta zeolite catalyst.
90.00
80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
0
50
100
150
200
250
300
Time / min
Fig. 7. Percentage conversion of ethyl cyanoacetate over the hydrotalcite catalyst as a function of time.
that showed a Mg–Sc–CO3 catalyst displays stronger base sites and
thus a higher activity resulting in greater conversions than Mg–
Mn–CO3 in the Knoevenagel condensation of ethyl cyanoacetate
and benzaldehyde [12].
Fig. 9 shows a comparison for the three different catalysts over
a period of 5 h, together with the blank run. It can be seen that the
mel-20-beta hybrid zeolite catalyst performed the best after five
hours with the order of reactivity, mel-20-beta > hydrotal-
cite > MgO/Al2O3 > un-catalysed. The same order was found for
the initial reaction rates as can be observed from Fig. 9. It can be
seen that the mel-20-beta zeolite has a significantly higher initial
reaction rate which is over three times higher than the hydrotal-
cite. This suggests that the varying basicity of the catalysts plays
a significant role in the kinetics of the reaction.
3.4. Variation of the mechanism depending on catalyst type
The different rates of reaction and end conversions of the Knoe-
venagel condensation suggest different reaction mechanisms for
Hydrotalcite
100.00
MgO/Al2O3
80.00
60.00
40.00
20.00
0.00
Mel-BEA
blank
100.00
80.00
60.00
40.00
20.00
0.00
0
50
100
150
200
250
300
-20.00
0
50
100
150
200
250
300
Time (min)
Time / min
Fig. 9. Comparison of percentage conversion of ethyl cyanoacetate over hydrotal-
cite, MgO/Al2O3, the mel-20-beta zeolite catalyst and the blank as a function of
time.
Fig. 8. Percentage conversion of ethyl cyanoacetate over MgO/Al2O3 as a function of
time.

18
A.C. Brooks et al. / Journal of Catalysis 285 (2012) 10–18
the formation of the ethyl trans-a-cyanocinnamate product. The
References
observation of increased performance for not fully exchanged zeo-
lite samples also suggests the necessity of some acid sites being
present that enhance the catalytic performance of the newly devel-
oped catalysts. Schemes 4 and 5 illustrate proposed mechanisms
that utilises acid (h-beta) and base (mel-100-beta) catalysts
respectively. The possible role of the melaminium cation as a basic
catalyst is shown in the upper part of Scheme 5. Although the med-
ium of pKb measurements are different to our reaction medium, it
can be noted that melamine has a first pKb of 9, and though the pKb
of its second protonation is not reported, a range of its deproto-
nated salts are easily prepared.
In our case (over mel-20-beta), a combination of these steps is
speculated to be operating (Scheme 6). Active methylene groups
upon deprotonation would react with the activated benzaldehyde,
leading to the synthesis of the condensation product and a mole-
cule of water. Indeed, a synergistic effect of acidic and basic sites
has also been suggested for Aldol condensations over mesoporous
materials with sulphonate and amine groups [14]. The sulphonate
group in this case was suggested to aid the reaction via deprotona-
tion–protonation of the acetone. This would be congruent with our
results as zeolite beta in its acidic form produced poor conversions.
Converting the majority of the acidic sites of the zeolite beta cata-
lyst to basic sites using 100 mol% melamine based on aluminium
content also produced a poor conversion. Besides a synergistic ef-
fect, the increased coverage could cause diffusion problems and
thereby the converted products and the reactants would have
greater difficulty exiting and entering the pore channels of the
catalyst.
[1] S. Feast, J.A. Lercher, in: H. Chon, S.I. Woo, S.E. Park (Eds.), Recent Advances
and New Horizons in Zeolite Science and Technology, Elsevier Science Publ. BV,
Amsterdam, 1996, pp. 363–412.
[2] K. Tanabe, W.F. Holderich, Appl. Catal. A – Gen. 181 (1999) 399–434.
[3] Y. Okamoto, M. Ogawa, A. Maezawa, T. Imanaka, J. Catal. 112 (1988) 427–
436.
[4] R.J. Davis, J. Catal. 216 (2003) 396–405.
[5] L.R.M. Martens, P.J. Grobet, P.A. Jacobs, Nature 315 (1985) 568–570.
[6] F. Winter, M. Wolters, A.J. van Dillen, K.P. de Jong, Appl. Catal. A – Gen. 307
(2006) 231–238.
[7] H. Mei, M. Hu, H.X. Ma, H.Q. Yao, J. Shen, Fuel Process. Technol. 88 (2007) 343–
348.
[8] J.H. Kwak, J. Szanyi, C.H.F. Peden, Catal. Today 89 (2004) 135–141.
[9] T. Seki, M. Onaka, J. Mol. Catal. A – Chem. 263 (2007) 115–120.
[10] A. Derrien, G. Renard, D. Brunel, Mesoporous Mol. Sieves (1998) 445–452.
[11] S. Jaenicke, G.K. Chuah, X.H. Lin, X.C. Hu, Microporous Mesoporous Mater. 35–
36 (2000) 143–153.
[12] X.H. Lin, G.K. Chuah, S. Jaenicke, J. Mol. Catal. A – Chem. 150 (1999) 287–
294.
[13] B.M. Choudary, M.L. Kantam, P. Sreekanth, T. Bandopadhyay, F. Figueras, A.
Tuel, J. Mol. Catal. A – Chem. 142 (1999) 361–365.
[14] K. Narasimharao, M. Hartmann, H.H. Thiel, S. Ernst, Microporous Mesoporous
Mater. 90 (2006) 377–383.
[15] J. Zhou, Z. Hua, J. Zhao, Z. Gao, S.Z. Zeng, J.L. Shi, J. Mater. Chem. 20 (2010)
6764–6771.
[16] L. Martins, D. Cardoso, Microporous Mesoporous Mater. 106 (2007) 8–16.
[17] M.J. Climent, A. Corma, S. Iborra, A. Velty, J. Mol. Catal. A – Chem. 182 (2002)
327–342.
[18] R.K. Zeidan, M.E. Davis, J. Catal. 247 (2007) 379–382.
[19] A. Corma, V. Fornes, R.M. Martinaranda, H. Garcia, J. Primo, Appl. Catal. 59
(1990) 237–248.
[20] L. Costa, G. Camino, J. Thermal Anal. Calorimetry 34 (1988) 423–429.
[21] CIPCS, Melamine, International Programme on Chemical Safety, 2005.
[22] S. Chen, Y. Yang, K. Zhang, J. Wang, Catal. Today 116 (2006) 2–5.
[23] J.M. Newsam, M.M.J. Treacy, W.T. Koetsier, C.B.D. Gruyter, Proc. Roy. Soc. Lond.
A Math. Phys. Sci. 420 (1988) 375–405.
Using our data and literature, we can infer that a synergistic ef-
fect may be operational between the melamine base sites and the
protonic acid sites. Scheme 6 outlines a likely mechanism, where
the reaction between the deprotonated cyanoester and the proton-
ated benzaldehyde ion is expected to increase the reactivity.
[24] M.M.J. Treacy, J.M. Newsam, Nature 332 (1988) 249–251.
[25] A. Hayashi, H. Nakayama, M. Tsuhako, Solid State Sci. 11 (2009) 1007–1015.
[26] M.L.M. Bonati, R.W. Joyner, M. Stockenhuber, Microporous Mesoporous Mater.
104 (2007) 217–224.
[27] L. Martins, K.M. Vieira, L.M. Rios, D. Cardoso, Catal. Today 133-135 (2008) 706–
710.
[28] L. Martins, W. Hoelderich, D. Cardoso, J. Catal. 258 (2008) 14–24.
[29] T. Hasegawa, C.K. Krishnan, M. Ogura, Microporous Mesoporous Mater. 132
(2010) 290–295.
4. Conclusions
[30] F.H. Allen, Acta Crystallogr. Sect. B – Struct. Sci. 58 (2002) 380–388.
[31] H.P. Jing, M. Strobele, M. Weisser, H.J. Meyer, Zeitschrift Fur Anorganische Und
Allgemeine Chemie 629 (2003) 368–370.
[32] J. Janczak, G.J. Perpetuo, Acta Crystallogr. Sect. C – Crystal Struct. Commun. 57
(2001) 1431–1433.
[33] K. Wijaya, O. Moers, D. Henschel, A. Blaschette, P.G. Jones, Zeitschrift Fur
Naturforschung (Section B – A J. Chem. Sci.) 59 (2004) 747–756.
[34] G.J. Perpetuo, J. Janczak, Acta Crystallogr. Sect. C – Crystal Struct. Commun. 62
(2006) O372–O375.
[35] A. Heine, K. Gloe, T. Doert, Zeitschrift Fur Anorganische Und Allgemeine
Chemie 634 (2008) 452–456.
Simple modification of NH4-beta zeolite with melamine leads to
a catalyst containing extremely stable cationic melamine species in
the cavities of the zeolite. This hybrid material is basic in nature. It
exhibits good performance in the Knoevenagel condensation of
benzaldehyde and ethyl cyanoacetate. The reaction rate, as well
as the conversion of 93% was either comparable or significantly im-
proved compared with other alkaline catalysts reported in the lit-
erature and to an acidic beta zeolite catalyst. We observed initial
high rates that were higher than the initial reaction rates shown
by the hydrotalcite and the MgO/Al2O3 catalysts. The activity in
the condensation reaction is dependent on the amount and nature
of acidic and basic sites present. The general stability of the cata-
lyst shows potential for application to industrial scale reactions
requiring mild basic and normally higher temperature conditions.
[36] A. Martin, A.A. Pinkerton, Acta Crystallogr. Sect. C – Crystal Struct. Commun. 51
(1995) 2174–2177.
[37] R.V. Siriwardane, M.S. Shen, E.P. Fisher, Energy Fuels 19 (2005) 1153–1159.
[38] M. Stockenhuber, J.A. Lercher, Microporous Mater. 3 (1995) 457–465.
[39] J. Howard, P.J. Lux, J. Yarwood, Zeolites 8 (1988) 427–431.
[40] S. Akyuz, T. Akyuz, J. Incl. Phenom. Macrocyc. Chem. 48 (2004) 75–80.
[41] S. Saravanamurugan, M. Palanichamy, M. Hartmann, V. Murugesan, Appl.
Catal. A – Gen. 298 (2006) 8–15.
[42] R.W. Joyner, A.D. Smith, M. Stockenhuber, M.W.E. van den Berg, Phys. Chem.
Chem. Phys. 6 (2004) 5435–5439.
[43] A.B. Wiles, D. Bozzuto, C.L. Cahill, R.D. Pike, Polyhedron 25 (2006) 776–782.
[44] C.J. Pouchert, The Aldrich Library of Infrared Spectra. Aldrich Chemical
Company, Milwaukee, WI, 1997, 5100.
[45] D. Lin-Vien, N.B. Colthup, W.G. Fateley, J.G. Graselli, The Handbook of Infrared
and Raman Characteristic Frequencies of Organic Molecules, Academic Press,
San Diego, CA, 1991. 503.
[46] J. Pironon, M. Pelletier, P. de Donato, R. Mosser-Ruck, Clay Minerals 38 (2003)
201–211.
[47] X.F. Zhang, E.S.M. Lai, R. Martin-Aranda, K.L. Yeung, Appl. Catal. A – Gen. 261
(2004) 109–118.
Acknowledgments
Financial support by Nottingham Trent University, UK and The
University of Newcastle, Australia is gratefully acknowledged.
The authors would like to thank Prof. B.Z. Dlugogorski, Prof. E.M.
Kennedy and Prof. R.W. Joyner for useful discussions. We also
acknowledge the use of CDS and the X-ray suite at UN and the Aus-
tralian synchrotron and the SRS for the use of the facilities to deter-
mine aluminium EXAFS of the zeolite beta.
[48] K. Komura, T. Kawamura, Y. Sugi, Catal. Commun. 8 (2007) 644–648.