Aldol condensation of furfural and acetone on zeolites

Article abstract of DOI:10.1016/j.cattod.2013.10.059

Zeolites of different structural types were used as catalysts for aldol condensation of furfural and acetone in batch reaction conditions at T = 20-100 C and time 0-24 h. To establish a relation between physico-chemical and catalytic properties of microporous materials, the samples were characterized by SEM, N₂ adsorption, FTIR and TGA. It was found that the acidic solids possessed appreciable activity in the reaction and resulted in a formation of products of aldehyde-ketone interaction. Nevertheless, furfural conversion decreased rapidly due to coke formation inside zeolite pores. Simultaneously with a general route of the reaction observed for basic catalysts, dimerization of the condensation product on acidic sites occurred. It was supposed that catalytic behavior of zeolites considerably affected by their both structural and textural properties. Experiments with re-used samples showed that zeolites totally restored their activity and selectivity after calcination at 530 C.

Full text of DOI:10.1016/j.cattod.2013.10.059

Catalysis Today 227 (2014) 154–162
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Aldol condensation of furfural and acetone on zeolites
Oleg Kikhtyanina, Vendula Kelbichováa, Dana Vitvarováb, Martin Kubu˚ ,
b
David Kubicˇkaa,∗
a Research Institute of Inorganic Chemistry, RENTECH-UniCRE, Chempark Litvínov, Záluzˇí–Litvínov, 436 70, Czech Republic
b J. Heyrovsky´ Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, v.v.i., Dolejsˇkova 3, 182 23 Prague, Czech Republic
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 26 July 2013
Received in revised form 13 October 2013
Accepted 18 October 2013
Available online 27 November 2013
Zeolites of different structural types were used as catalysts for aldol condensation of furfural and acetone
in batch reaction conditions at T = 20–100 ◦C and time 0–24 h. To establish a relation between physico-
chemical and catalytic properties of microporous materials, the samples were characterized by SEM,
N2 adsorption, FTIR and TGA. It was found that the acidic solids possessed appreciable activity in the
reaction and resulted in a formation of products of aldehyde–ketone interaction. Nevertheless, furfural
conversion decreased rapidly due to coke formation inside zeolite pores. Simultaneously with a general
route of the reaction observed for basic catalysts, dimerization of the condensation product on acidic sites
occurred. It was supposed that catalytic behavior of zeolites considerably affected by their both structural
and textural properties. Experiments with re-used samples showed that zeolites totally restored their
activity and selectivity after calcination at 530 ◦C.
Keywords:
Aldol condensation
Oligomerization
Zeolites
Acidity
© 2013 Elsevier B.V. All rights reserved.
Coke formation
Regeneration
1. Introduction
Aldol condensation of aldehydes and ketones is a well-known
reaction of organic synthesis and it proceeds in the presence of
In conditions of unpredictable prices for conventional hydrocar-
bon sources, it is of a great importance to develop new technologies
based on non-traditional, renewable energy supplies. At present,
significant research attention is focused on the transformation of
waste biomass into fuels and valuable chemicals [1,2]. The primary
transformation of biomass can proceed by different routes result-
ing in a formation of bio-oil (by liquefaction and/or pyrolysis), sugar
solutions, etc. [1–6]. Acetone, which is obtained as a product of
acetic acid decomposition (ketonization), and furanic compounds
that are afforded by sugars hydrolysis are among the final prod-
ucts of the successive transformation steps involved in conversion
of the starting complex biomass mixtures. Furanic compounds and
acetone, being aldehydes and a ketone by chemical origin, may fur-
ther react with each other via aldol condensation in the presence
of the corresponding catalyst [7–9]. In this respect, aldol conden-
sation seems to be a useful tool to increase a length of carbon chain
and to produce heavier reaction products from comparatively light
initial reagents. The aldol condensation step is of particular impor-
tance in upgrading of bio-oils since these light molecules would
be otherwise converted into low-value gaseous products during
bio-oil deoxygenation that is necessary to improve its stability and
enhance its fuel properties [10].
catalysts with either basic or acidic properties. Most often basic
catalysts are used for this reaction because they possess high
activity in conversion of reagents and high selectivity toward
the desired reaction products (Scheme 1). Sodium hydroxide is
a recognized catalyst for aldol condensation, which permits to
carry out the reaction even at room temperature [8]. Neverthe-
less, high corrosivity of this compound and the increasing demands
for environmental friendliness of new technologies result in a
necessity of using heterogeneous catalytic systems with similar
basic properties. Catalysts derived from hydrotalcite-like (HTC)
materials as well as other mixed oxide systems were found to
be good heterogeneous catalysts with basic properties, which
were successively employed as catalysts for aldol condensation
[11–14]. It is known that re-hydrated HTC materials allow car-
rying out aldol condensation between furfural and acetone with
high furfural conversion already at 50–60 ◦C [13]. Nevertheless, a
number of published works dealing with the properties of HTC
materials indicate significant disadvantages of these materials;
in particular high sensitivity to the ambient CO2 and problems
with re-using the HTC-based catalysts in successive catalytic
runs [11,15,16].
Solid catalysts with acidic properties are used seldom as
catalysts for aldol condensation in comparison with basic het-
erogeneous catalysts. Zeolites, being materials with pronounced
and controllable acidic properties, seem to be suitable mate-
rials for this purpose. Indeed, it is known that acidic zeolites
∗
Corresponding author. Tel.: +420 476163735.
E-mail address: david.kubicka@vuanch.cz (D. Kubicˇka).
0920-5861/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cattod.2013.10.059

O. Kikhtyanin et al. / Catalysis Today 227 (2014) 154–162
155
Scheme 1. Aldol condensation between furfural and acetone over acidic or basic catalysts.
of different structural types can catalyze condensation of vari-
ous aldehydes and ketones [17–19]. However, the examples of
using zeolites as catalysts for aldol condensation are still few.
Therefore, further investigation to gain deeper understanding of
the suitability of these microporous materials as catalysts for
reactions of different biomass-derived aldehydes and ketones
are required to promote the development of efficient biomass
transformation routes into fuel components and valuable chemi-
cals.
133 Pa, 1.33 kPa and 133 kPa ranges. The zeolite samples were
outgassed under turbomolecular pump vacuum before start-
ing the adsorption experiments. The zeolite samples were
outgassed at 110 ◦C (heating rate 1 ◦C/min starting from the
ambient temperature) until the residual pressure of 0.5 Pa was
obtained. After further heating at 110 ◦C for 1 h the tempera-
ture was increased until the temperature 300 ◦C (heating rate
1 ◦C/min) was achieved. This temperature was maintained for
6 h.
The present work aims at understanding of the relationship
between the properties of zeolites of different structural type and
their catalytic performance in aldol condensation of furfural and
acetone. Both activity and selectivity of these materials are corre-
lated with their physico-chemical properties.
Acidic properties of zeolites were examined by IR spectroscopy
of adsorbed pyridine on Nicolet 6700 spectrometer equipped
with MCT detector. Pyridine adsorption proceeded at 150 ◦C for
20 min at partial pressure 3 Torr, followed by 20 min evacuation
at 150 ◦C. The concentrations of Brønsted and Lewis acid sites
were calculated from integral intensities of the individual bands
characteristic of pyridine adsorbed on Brønsted acid sites (BAS,
at 1545 cm−1) and on Lewis acid sites (LAS, at 1455 cm−1) and
molar absorption coefficients of ε(B) = 1.67 0.1 cm ␮mol−1 and
ε(L) = 2.22 0.1 cm ␮mol−1, respectively [20]. The spectra were
recorded with a resolution of 4 cm−1 by collection of 128 scans for a
single spectrum. Prior to the adsorption of pyridine, self-supporting
wafers of samples were activated in situ by overnight evacuation
at temperature 450 ◦C.
2. Experimental
2.1. Supply and preparation of catalysts
Zeolite samples for the study (NH4-ZSM-5(23), NH4-ZSM-
5(50), NH4-BEA(25), NH4-BEA(38), NH4-MOR(20), HSDUSY(80),
HSUSY(5)) were provided by Zeolyst International. The numbers
in parentheses designate SiO2/Al2O3 molar ratio of the samples as
given by the manufacturer. The transformation of NH4-form of zeo-
lites to the corresponding H-form was carried out by calcination in
air at 500 ◦C for 4–6 h. The effect of chemical composition of the
samples on their physico-chemical and catalytic properties was
investigated by using materials of the same structural type with
different SiO2/Al2O3 molar ratio.
2.3. Catalysis
Properties of zeolite samples were investigated in aldol con-
densation of furfural and acetone. The catalytic experiments were
carried out in a 100 ml stirred batch reactor (a glass flask reactor
with magnetic stirring for the experiments at 20 and 60 ◦C, Parr
stirred autoclave for the experiments at 100 ◦C). Before the start
of the catalytic runs at 20 and 60 ◦C, 2 g of catalyst was mixed
together with a stirred mixture of 39.5 g acetone and 6.5 g of fur-
fural (acetone/furfural molar ratio 10/1, pre-heated to the desired
reaction temperature, 20 or 60 ◦C) and kept at the reaction tempera-
ture for 2–24 h under intensive stirring. For autoclave experiments,
2 g of catalyst was mixed together with the reaction mixture of
the above composition and loaded into the autoclave. After ini-
tiation of the heating the desired temperature was achieved in
2.2. Physico-chemical characterization
The shape and size of the zeolite crystals were determined
by scanning electron microscopy (SEM, Jeol, JSM-5500LV). The
textural properties of the samples were determined by nitro-
gen physisorption at 77 K using an ASAP 2020 (Micromeritics)
static volumetric apparatus. In order to attain sufficient accu-
racy in the accumulation of the adsorption data, the ASAP
2020 was equipped with pressure transducers covering the

156
O. Kikhtyanin et al. / Catalysis Today 227 (2014) 154–162
∼60 min, and the autoclave was kept at T = 100 ◦C for additional
0, 2 or 8 h.
3. Results and discussion
Analysis of the reaction products was performed on an Agilent
7890A GC unit equipped with a flame ionization detector using a
HP-5 capillary column (30 m/0.32 ID/0.25 ␮m). The obtained prod-
ucts were identified based on the standard reference compounds
as well as additional GC-MS analyses.
3.1. Physico-chemical characterization
The performed XRD analysis confirmed that all the samples were
phase-pure zeolites of the declared structure with a high degree
of crystallinity. Fig. 1 shows SEM images of the zeolite samples. It
Fig. 1. SEM images of zeolites samples. (A) NH4-ZSM-5(23), (B) NH4-ZSM-5(50), (C) NH4-BEA(25), (D) NH4-BEA(38), (E) HSDUSY(80), (F) HSUSY(5), (G) NH4-MOR(20).

O. Kikhtyanin et al. / Catalysis Today 227 (2014) 154–162
157
Table 1
Textural properties of zeolite samples.
Sample
BET (m2/g)
Sext (m2/g)
Vmic (cm3/g)
Vmeso (cm3/g)
Vtot (cm3/g)
DBJH (Å)
HZSM-5(23)
HZSM-5(50)
HBEA(25)
HBEA(38)
HSDUSY(80)
HSUSY(5)
378
398
608
617
840
588
471
99
133
222
134
294
94
0.126
0.118
0.174
0.218
0.246
0.226
0.190
0.055
0.102
0.851
0.036
0.210
0.166
0.015
0.210
0.255
1.017
0.334
0.536
0.388
0.269
10.9
10.0
9.2
–
13.6
7.4
–
HMOR(20)
55
BET, surface area (from BET method); Vmic, micropore volume (from t-plot method); Vmeso, mesopore volume (from BJH method); Vint, interparticle space volume; Vtot, total
pore volume at p/p0 = 0.97; D, average pore diameter (from BJH method, for mesoporous samples).
can be seen that the size and shape of the crystals of MFI zeolites
differ between the samples with different Al content (Fig. 1A and B).
NH4-ZSM-5(23) sample is represented by well-defined individual
crystals of 0.5–1 ␮m. In contrast, the crystals of NH4-ZSM-5(23) do
not possess a certain shape and their size varies from 100 to 300 nm.
The crystals of two samples with BEA structure are also different
(Fig. 1C and D). NH4-BEA(25) consists of very large agglomeration
of small crystals with the size smaller than 100 nm. SEM image
of NH4-BEA(38) indicates the presence of spherical particles in a
wide range of 0.2–1 ␮m. Crystal morphology of two FAU samples
is similar: they consist of crystals with dimensions of 0.2–1 ␮m
(Fig. 1E and F). The size of crystals of MOR zeolite lies in the same
wide range of 0.2–1 ␮m (Fig. 1G).
The adsorption characteristics of zeolites of different structural
types are presented in Table 1. Obviously, BET surface is determined
by a structural type of the zeolite material. This value is the highest
for the three-dimensional wide-pore BEA (608–617 m2/g) and FAU
(588–840 m2/g) materials, while the lowest value is observed for
medium-pore MFI (378–398 m2/g). MOR zeolite, which belongs to
one-dimensional wide-pore microporous systems, takes an inter-
mediate position. Both BEA samples have similar BET surface
values, despite the fact that, according to SEM, their crystal mor-
phology varies significantly. This can be explained by assuming that
the particles of H-BEA(38) zeolite with a visual size of 0.2–1 ␮m
consist in fact of very small crystals similar to those of H-BEA(25)
zeolite. The differences in the micropore volume of the samples
are similar to those observed for the BET surface: the highest value
Vmic is observed for BEA and FAU zeolites, and the lowest - for MFI.
Besides, the results presented in Table 1 indicate that mesoporosity
of the samples varies to a great extent. For example, the mesopore
volume is in the range 0.055–0.1 cm3/g for MFI samples and in the
range 0.166–0.21 cm3/g for FAU samples. In the case of BEA zeo-
lites the difference in mesoporosity is more pronounced: it reaches
0.85 cm3/g for H-BEA(25), but it is almost absent for H-BEA(38)
(Figs. 2 and 3).
The concentration of strong BAS and LAS present in the samples
was determined by FTIR of adsorbed pyridine (Table 2). Both the
amount of these sites and BAS/LAS ratio depend to a large extent
on the Al content in the samples. For zeolites of the same structural
type an increase in Al content results in an increased concentra-
tion of LAS and, as a result, in a decreased BAS/LAS ratio. For MFI
materials an increase of the SiO2/Al2O3 ratio leads to a decreased
180
700
1
2
600
160
500
B
140
400
300
A
120
A
B
100
80
200
100
0
0,2
0,4
0,6
0,8
1
0
0,2
4
0,4
0,6
0,8
1
Relative Pressure (p/po)
Relative Pressure (p/p°)
200
160
120
340
300
260
220
180
140
3
A
B
0
0,2
0,4
0,6
0,8
1
0
0,2
0,4
0,6
0,8
1
Relative Pressure (p/p°)
Relative Pressure (p/p°)
Fig. 2. Isotherms of zeolite samples. 1 – HZSM-5 (A – HZSM-5(23), B – HZSM-5(50)), 2 – HBEA (A – HBEA(25), B–HBEA(38)), 3 – FAU (A – HSDUSY(80), B – HSUSY (5)), 4 –
HMOR(20).

158
O. Kikhtyanin et al. / Catalysis Today 227 (2014) 154–162
Fig. 3. IR spectra in the OH vibration region (1) and spectra after pyridine adsorption (2). (A) HMOR(20), (B) HZSM-5(23), (C) HZSM-5(50), (D) HBEA(25), (E) HBEA(38), (F)
HSDUSY(80), (G) HSUSY (5).
Table 2
It is evident that crystal structure of zeolites introduces consider-
able influence on their behavior in the reaction. Medium pore MFI
Acidic properties of zeolite samples obtained by FTIR method.
zeolites and wide-pore mono-dimensional MOR zeolite possess
low activity, and furfural conversion over these catalysts does not
exceed 10% even at T = 100 ◦C (Table 3). On the other hand, the activ-
ity of wide-pore zeolites with three-dimensional crystal framework
increases when they are transformed into H-form, and furfural
conversion over BEA zeolites achieves 30% at T = 100 ◦C after 2 h
of the reaction. It can be assumed that in addition to the crystal
structure additional characteristics, for example, textural and acid-
ity properties, influence the catalytic properties of the molecular
sieves. However, taking together FTIR data with catalytic results
it becomes obvious that not only the acidic properties of zeolites
introduce a decisive influence on their catalytic activity in the reac-
tion. A comparison of acidic properties of the two iso-structural
HBEA samples indicates that BAS concentration in these materials is
similar, and LAS concentration is higher for the sample with higher
Al content, namely HBEA(25). At the same time, their activity in
aldol condensation is approximately the same. It leads to the con-
clusion that the reaction between furfural and acetone over zeolites
proceeds with participation of BAS rather than LAS. Nonetheless,
comparison of two FAU samples shows that HSDUSY(80) with
lower BAS and, in particular, LAS, concentrations is more active
in the reaction than HSUSY(5). The same tendency also occurs in
the case of two MFI zeolites: activity of HZSM-5(50) with lower
acidity is higher compared with that of HZSM-5(23). These depen-
dences lead to the conclusion that the acidity of the zeolite is not
the single factor that determines the catalytic performance of the
investigated zeolites in aldol condensation.
Study of textural characteristics of zeolites and their SEM data
show that materials with smaller crystal sizes possess larger values
of BET surface area and external surface area. Obviously, the larger
the external surface, the greater is the amount of pore openings
available for the access of the reactant molecules into the internal
pore system. Indeed, when correlating the activity of the catalysts
with their textural characteristics, it is seen that the materials with
larger BET surface values exhibit higher activity in the reaction.
Summarizing the results, it can be concluded that the activity of
zeolites in aldol condensation of furfural and acetone increases
when wide-pore three-dimensional microporous materials with
small crystal size are used.
Sample
CLAS (mmol/g)
CBAS (mmol/g)
CLAS/CBAS
HZSM-5(23)
HZSM-5(50)
HBEA(25)
HBEA(38)
HSDUSY(80)
HSUSY(5)
0.220
0.049
0.255
0.179
0.068
0.223
0.241
0.353
0.183
0.138
0.142
0.104
0.127
0.282
1.6
3.73
0.54
0.79
1.53
0.57
1.17
HMOR(20)
concentration of BAS, but for BEA and FAU zeolites the same trend
is not obvious.
3.2. Catalysis
3.2.1. The catalytic performance of the zeolites
Table 3 contains the results on furfural conversion over zeo-
lites with different structural types and nature of the compensating
cations (NH4 and H+). It can be clearly seen that the properties
+
of the investigated zeolites substantially influence their activity in
aldol condensation. Regardless of the type of crystal framework, all
the samples in the NH4-form possess very low activity in the reac-
tion. In contrast, the application of zeolites in the H-form results
in a noticeable increase of furfural conversion, especially at high
reaction temperature. The difference in the catalytic performance
between the samples in NH4- and H-forms unambiguously proves
that aldol condensation of furfural and acetone takes place with the
participation of acidic sites of zeolites. A correlation of the catalytic
performance of the zeolite samples with their physico-chemical
characteristics permits to evaluate the key factors determining the
activity of zeolites in aldol condensation of furfural and acetone.
Table 3
Furfural conversion (%) at different reaction temperatures (after 2 h of the reaction).
Sample
Reaction temperature (◦C)
20
60
0
1.5
0
3.1
0
1.1
0
19.7
0
22.6
10.4
1.3
100
NH4-ZSM-5(23)
HZSM-5(23)
NH4-ZSM-5(50)
HZSM-5(50)
NH4-MOR
HMOR(20)
NH4-BEA(25)
HBEA(25)
NH4-BEA(38)
HBEA(38)
HSDUSY(80)
HSUSY(5)
0
0
0
0
0
0
0
2.8
0
0
6.4
0.7
9.5
0.7
6.8
7.1
38.5
6.2
Furfural conversion over different zeolites in dependence on
reaction time at T = 60 and 100 ◦C is shown on Fig. 4. It is seen
that furfural conversion increases very fast during the initial
period of the reaction, and after approximately 2 h the catalysts
achieve a quasi-stationary state, i.e. the conversion of furfural
increases only very slightly. Among the wide-pore materials,
BEA zeolites are the most active microporous materials at any
6.5
1.6
0
29.0
21.8
12.0

O. Kikhtyanin et al. / Catalysis Today 227 (2014) 154–162
159
40
35
30
25
20
15
10
5
55
B
A
50
45
40
35
30
25
20
15
10
5
Heating
0
0
0
4
8
12
16
20
24
0
1
2
3
4
5
6
7
8
Reaction Time, h
Reaction Time, h
Fig. 4. Conversion of furfural over different zeolites in H-form at 60 ◦C (A) and 100 ◦C (B). (ꢀ) HMOR(20), (ꢁ) HZSM-5(28), (ꢂ) HZSM-5(50), (ꢃ) HSUSY(5), (ꢄ) HSDUSY(80),
(•) HBEA(25), (ꢅ) HBEA(38).
reaction time as compared to FAU zeolites. An evaluation of
the initial and quasi-stationary reaction rates can be performed
using the angles of the slopes of the lines derived from the
kinetic curves in the range 0–2 h and 4–24 h, respectively. The
calculations show that for BEA and FAU zeolites the initial and
quasi-stationary reaction rates differ by more than an order
of magnitude (Wini/Wquasi = tg␣0−2/tg␣4−24 = 30 − 40), which
indicates severe deactivation.
given reaction conditions the subsequent condensation of the sec-
ond furfural molecule with FAc hardly occurs. An explanation might
be the low concentration of furfural, or more plausibly, the limited
dimensions of the pores that hinder the formation and diffusion of
the bulky F2Ac product.
Simultaneously, the formation of a heavier reaction product
than F2Ac is observed (Fig. 4). GC–MS analysis of this compound
proves its molecular weight to be 272. The molecular mass cor-
responds to a dimer of FAc and it can thus be assumed that this
compound is a result of dimerization of olefinic FAc molecule on
acid sites of zeolites (Scheme 2). However, at present it remains
still unclear at which double bond the dimerization occurs and
additional experiments will be carried out to clarify this issue. The
content of this dimer (FAc)2 also increases with the growth of the
furfural conversion. Fig. 5 also shows that yield of this dimer com-
pound is larger when using HSDUSY zeolite as a catalyst compared
with HBEA(25). Taking into account the data on textural and acidic
properties of these zeolites (Tables 1 and 2) it may be assumed
that the formation of this dimer readily occurs inside microporous
volume of zeolites, and large cages of FAU structure are the most
suitable for this reaction route. Similar composition of reaction
products and their dependence on furfural conversion are observed
when the reaction temperature is increased to 100 ◦C (Fig. 5B).
The formation of this dimer compound was not yet reported in
literature dealing with aldol condensation on HTC materials and
other basic catalysts. By comparing the composition of the reac-
tion products which is observed on basic materials (double oxides)
and on acidic materials (zeolites) it could be concluded that for
The composition of the reaction products obtained on zeolites
with different structure types substantially depends on the activity
of the catalysts or more specifically on the level of furfural conver-
sion. For the samples with low activity, i.e. MFI and MOR zeolites,
FAc is as the major reaction product with selectivity ≥95% in the
range of reaction temperatures 20–100 ◦C. The single by-product
in this case is the corresponding alcohol, which is in line with the
consecutive reaction mechanism and proves its fast dehydration to
FAc on acidic sites of zeolites (Scheme 1). The use of the wide-pore
three-dimensional zeolites results in an increased furfural conver-
sion, and, as a result, the composition of the products formed in
aldol condensation also changes. The yield of the reaction products
as a function of furfural conversion observed over HBEA(25) and
HSDUSY(80) catalysts at T = 60 ◦C is plotted in Fig. 5A. FAc is the
major compound formed in the reaction, and its yield increases
linearly as the furfural conversion increases. The corresponding
alcohol is absent among the reaction products due to its total and
fast dehydration at this reaction temperature. Independently on
the furfural conversion, the content of F2Ac, i.e. the subsequent
condensation product (Scheme 1), is very low. This proves that at
30
40
A
B
35
30
25
20
15
10
5
25
20
15
10
5
0
0
0
5
10
15
20
25
30
35
40
45
50
0
5
10
15
20
25
30
35
Furfural Conversion, %
Furfural Conversion, %
Fig. 5. Yield of reaction products formed in aldol condensation of furfural and acetone over HBEA(25) (ꢀ,•,ꢁ) and HSDUSY(80) (ꢆ, ꢀ, ꢀ) at 60 ◦C (A) and 100 ◦C (B). (ꢀ) FAc,
(•) F2Ac, (ꢁ) (FAc)2.

160
O. Kikhtyanin et al. / Catalysis Today 227 (2014) 154–162
Scheme 2. Formation of dimer product from FAc compound.
both types of catalysts FAc is the primary general product of aldol
condensation which is formed by dehydration of the correspond-
ing alcohol. A further transformation of the formed FAc compound
proceeds by different routes on either basic or acid catalysts. The
consecutive aldol condensation of FAc to F2Ac takes readily place
over HTC, i.e. basic catalysts, while zeolites favor mostly dimeriza-
tion of FAc to (FAc)2 due to the presence of unsaturated bonds in
FAc which can readily interact with Brønsted acid sites.
The position and intensity of the DTG peaks located in HT region
depend both on the structural type of the zeolite samples and
on the conditions of the thermal analysis. Heating of the spent
HBEA(25) sample in nitrogen atmosphere results in appearance of a
broad peak with maximum at 400 ◦C and weight loss of 5.3%. If the
experiment is performed in oxidative atmosphere, two HT peaks
are observed: at 380 ◦C with a weight loss of 4.1% and at 459 ◦C
with weight loss of 6.5% (Fig. 6). It may be concluded that peaks
at 380–400 ◦C correspond to desorption of heavy reaction prod-
ucts from the zeolite pores while the latter peak is attributed to
the oxidative destruction of the non-desorbed compounds, which
cannot escape from the internal zeolite volume. Fig. 6 also shows
that DTG profiles are nearly identical for all investigated zeolites
and the findings can be summarized as follows: carbonaceous
deposits formed during aldol condensation of furfural and acetone
are removed from the internal volume of microporous materials
in two stages. At 300–360 ◦C occurs either removal of light reac-
tion products or thermal degradation of heavier substances. At
450–510 ◦C takes place burning of the residual coke compounds
in the oxidative atmosphere.
3.2.2. TGA of the spent catalysts
As shown by the catalytic tests, the activity of zeolites in aldol
condensation of furfural and acetone is high at the beginning of
the reaction but rapidly decreases with the increasing time of the
experiment. The samples recovered from the reactor are black in
color suggesting a big amount of coke deposits in them. The cat-
alysts after reaction were thus abundantly washed with ethanol
to remove any organics from external surface of the zeolite crys-
tals, and the content of the carbonaceous deposits inside micropore
volume was investigated by TGA (Fig. 6).
Two regions of weight loss, low-temperature (LT) and high-
temperature (HT), are clearly found on the curves. Regardless of
the atmosphere in which the experiments have been carried out the
maximum of the LT peak is observed in the range of 57–78 ◦C, which
corresponds to the removal of residual light compounds from the
samples. The weight loss in this region is in the range 3.2–6.5%
(Table 4).
3.2.3. Catalytic properties of re-used samples
It is known that oxidative treatment does not result in a
total restoration of the catalytic activity of HTC materials in aldol
condensation [13]. This disadvantage substantially reduces the
prospects of these materials as catalysts for practical use. Unlike
Fig. 6. DTG profiles obtained in inert (dotted line) and oxidative (solid line) ambience for HBEA(25) (A), HSDUSY(80) (B) and HZSM-5(23) (C).

O. Kikhtyanin et al. / Catalysis Today 227 (2014) 154–162
161
Table 4
Results of TGA for zeolite samples carried out in inert (nitrogen) and oxidative (oxygen) atmospheres.
Sample
Conditions of TGA experiments
N2
Temperature region (◦C)
Maximum value on DTG curve (◦C)
Weight loss (%)
HBEA(25)
20–200
200–600
20–200
200–600
20–200
70
400
72
380; 459
60;
142
301
57; 140
318; 510
78
6.5
5.3
6.2
4.1; 6.2
3.3;
0.7
6.4
3.2; 0.5
4.5; 6.7
3.5
O2
N2
HSDUSY(80)
HZSM-5(23)
200–600
20–200
200–600
20–200
200–600
20–200
200–600
O2
N2
O2
362
78
338; 477
8.1
3.6
3.2; 6.7
Table 5
Catalytic properties of the re-used catalysts after washing with ethanol and calcination at 530 ◦C in comparison with the fresh HBEA(25) sample. Treac. = 100 ◦C (except for *,
Treac. = 60 ◦C), time of the reaction – 2 h.
Sample
Furfural conversion (%)
Selectivity (%)
FAc
F2Ac
(FAc)2
HBEA(25), 1st run
38.5
20.2
40.2
19.8
79.5
88.5
71.8
80.2
3.7
4.6
3.6
0
16.8
6.9
24.6
19.8
HBEA(25), 2nd run (washed with ethanol)
HBEA(25), 2nd run (calcined at 530 ◦C)
*HBEA(25), T = 60 ◦C (conversion ca. 20%)
HTC, most zeolites are known to be very resistant to oxidative
regeneration up to T = 500–550 ◦C and totally restore their proper-
ties during their re-using as catalysts in different organic reactions.
It is seen from Table 5 that washing of HBEA(25) with an excess
of ethanol does not restore the initial activity of HBEA(25) com-
pletely, and furfural conversion over the washed sample attains
only 20% comparing with 39% for the fresh catalyst. Additionally,
a change in the composition of the reaction products formed on
the washed sample is also observed: the selectivity toward the
first product of aldol condensation, FAc, increases and the selec-
tivity toward dimer product, (FAc)2, correspondingly, decreases.
Provided that the microporous volume of this zeolite after reaction
and washing is still filled with carbonaceous deposits, the mod-
erate decrease in conversion of furfural for the washed sample
proves that aldol condensation of furfural and acetone can pro-
ceed with participation of both internal and external acid sites of
HBEA(25). At the same time, a significant decrease of the selectivity
toward (FAc)2 indicates that this product predominantly is formed
inside micropore volume of HBEA(25) where a close interaction
of two FAc molecules can be ensured. Moreover, a comparison of
the reaction products observed at equal furfural conversion (ca.
20%) for two different samples HBEA(25), i.e. for the coked one at
100 ◦C and washed and for the fresh one at 60 ◦C also reveal that
(FAc)2 is formed independently on the conversion level if the micro-
pores are free from any deposits. After calcination of the coked
zeolite at 530 ◦C both furfural conversion and the composition of
the reaction products attain values observed for the fresh sample
(Table 1).
4. Conclusion
The presented results show that the acidic zeolites possess
rather good activity in aldol condensation of furfural and acetone
and their catalytic properties are determined by the structural type,
as well as by their textural and acidic characteristics. The highest
conversion of furfural is observed when using wide pore zeo-
lites with a three-dimensional crystalline framework. The obtained
results suggest that Brønsted acid sites are essential in the trans-
formation. In the presence of a solid acidic catalyst the reaction
between furfural and acetone results in the formation of FAc, which
is also the primary product of the aldol condensation in the pres-
ence of basic catalysts. However, the following transformation of
FAc proceeds by different reaction routes. Unlike basic materials,
the properties of the zeolites contribute to the formation of (FAc)2
compound which is formed via dimerization of the olefinic FAc on
the acid sites. During the reaction, the activity of the investigated
zeolites decreases due to the formation of carbonaceous deposits
inside their micropores. Nevertheless, their catalytic properties
in aldol condensation of furfural and acetone can be completely
restored by calcination at temperatures above 500 ◦C. The objec-
tives for the future investigations should be focused on an increase
of the activity of zeolites in the reaction and on improvement of
their stability by optimization of both physico-chemical properties
of the microporous materials and the reaction conditions.
Acknowledgements
On the other hand, the selectivity toward F2Ac is higher over
the coked HBEA(25) i.e. over the catalyst that has some of the acid
sites blocked, as compared with the fresh one tested at 60 ◦C at
equal furfural conversion of 20%. The observed difference in selec-
tivity of the catalysts allows concluding that the formation of F2Ac
takes place easily either with the participation of the active sites
located on the external surface of the zeolite crystals or with those
having lower acidic strength. Thus, the optimization of the physico-
chemical characteristics of the zeolite materials, including their
textural and acidic properties, could be a key factor for directing
the interaction between furfural and acetone to the desired reaction
route.
The authors thank the Czech Science Foundation for the
financial support (Center of Excellence – P106/12/G015). The
project (P106/12/G015) is being carried out in the UniCRE cen-
ter (CZ.1.05/2.1.00/03.0071) in the Research Institute of Inorganic
Chemistry.
References
[1] G. Huber, S. Iborra, A. Corma, Chem. Rev. 106 (9) (2006) 4044–4098.
ˇ
[2] D. Kubicˇka, I. Kubicˇková, J. Cejka, Catal. Rev. Sci. Eng. 55 (2013) 1–78.
[3] A.V. Bridgwater, Biomass Bioenerg 38 (2012) 68–94.
[4] D.L. Klass, Biomass for Renewable Energy, Fuels and Chemicals, Academic Press,
San Diego, 1998.

162
O. Kikhtyanin et al. / Catalysis Today 227 (2014) 154–162
[5] G.W. Huber, A. Corma, Angew. Chem. Int. Ed. 46 (2007) 7184–7201.
[13] L. Faba, E. Diaz, S. Ordõnez, Appl. Catal. B: Environ. 113–114 (2012) 201–211.
[14] M.J. Climent, A. Corma, S. Iborra, A. Velty, Green Chem. 4 (2002) 474–480.
[15] Chunli Xu, Yuan Gao, Xihong Liu, Ruirui Xin, Zhen Wang, RSC Adv. 3 (2013)
793–801.
[6] Y. Sun, J. Cheng, Bioresour. Technol. 83 (2002) 1–11.
[7] G.W. Huber, J.N. Chheda, C.J. Barret, J.A. Dumesic, Science 308 (2005)
1446–1450.
[8] R.M. West, Z.Y. Liu, M. Peter, C.A. Gärtner, J.A. Dumesic, J. Mol. Catal. A: Chem.
296 (2008) 18–27.
[16] S. Abelló, D. Vijaya-Shankar, J. Pérez-Ramírez, Appl. Catal. A: Gen. 342 (2008)
119–125.
[9] A.V. Rong Xing, H. Subrahmanyam, Olcay, G.P. Wei Qi, H. van Walsum, G.W.
Pendse, Huber, Green Chem. 12 (2010) 1933–1946.
[17] E. Dumitriu, V. Hulea, I. Fechete, A. Auroux, J.-F. Lacaze, C. Guimon, Microporous
Mesopororous Mater. 43 (2001) 341–359.
[10] T.N. Pham, D. Shi, D.E. Resasco, Appl. Catal. B: Environ. (2013),
http://dx.doi.org/10.1016/j.apcatb.2013.01.002.
[11] D.P. Debecker, E.M. Gaigneaux, G. Busta, Chem. Eur. J. 15 (2009) 3920–3935.
[12] I. Sádaba, M. Ojeda, R. Mariscal, R. Richards, M. López Granados, Catal. Today
167 (2011) 77–83.
[18] A. Ungureanu, S. Royer, T.V. Hoang, D. Trong On, E. Dumitriu, S. Kaliaguine,
Microporous Mesopororous Mater. 84 (2005) 283–296.
[19] T. Komatsu, M. Mitsuhashi, T. Yashima, Stud. Surf. Sci. Catal. 142 (2002)
667–674.
[20] C.A. Emeis, J. Catal. 141 (1993) 347–354.