ˇZeljka kesi´ca, ivana luki´ca, dragana brki´ca, jelena rogana, miodrag zduji´cb, hui liuc, dejan skala

1. Applied Catalysis A: General 427–428 (2012) 58–65
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Mechanochemical preparation and characterization of CaO·ZnO used as catalyst
for biodiesel synthesis
ˇZeljka Kesi´ca
, Ivana Luki´ca
, Dragana Brki´ca
, Jelena Rogana
, Miodrag Zduji´cb
, Hui Liuc
, Dejan Skalaa,∗
a
University of Belgrade, Faculty of Technology and Metallurgy, Karnegijeva 4, 11000 Belgrade, Serbia
b
Institute of Technical Sciences of the Serbian Academy of Sciences and Arts, Knez Mihailova 35, 11000 Belgrade, Serbia
c
University of Geosciences, School of Environmental Studies, Wuhan 430074, PR China
a r t i c l e i n f o
Article history:
Received 19 December 2011
Received in revised form 19 March 2012
Accepted 22 March 2012
Available online 1 April 2012
Keywords:
Biodiesel
Heterogeneous catalyst
Methanolysis
CaO
ZnO
a b s t r a c t
In this study, the synthesis of biodiesel or fatty acid methyl esters (FAME) from sunflower oil and methanol
using CaO·ZnO catalyst was investigated. Catalyst was synthesized by ball milling of Ca(OH)2 and ZnO
powder mixture with the addition of water (BMH), as well as solely by ball milling of mentioned pow-
ders (BM) and subsequent calcination at 700 ◦
C in air atmosphere. For comparison, the CaO·ZnO mixed
oxide was also prepared using usual coprecipitation procedure (CP) followed by calcination at 700 ◦
C of
the formed calcium zinc hydroxide hydrate. The BMH, BM and CP catalysts were characterized by X-ray
diffraction (XRD), thermogravimetric analysis (TGA), infrared spectroscopy (FTIR), particle size distribu-
tion measurement and scanning electron microscopy (SEM and SEM-EDS). In addition, specific surface
area (BET), solubility in methanol at 60 ◦
C and alkalinity (Hammett indicator method) were also deter-
mined. The activity of BMH, BM and CP catalysts for biodiesel synthesis were tested at 60 ◦
C and 1 bar,
using molar ratio of sunflower oil to methanol of 1:10 and with 2 wt% of catalyst based on oil weight.
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
Due to the limitation of fossil fuel and its toxicity today’s
research is directed towards alternative renewable sources.
Biodiesel is non-toxic and biodegradable renewable fuel derived
from vegetable oils, animal fats or used cooking oils. Biodiesel
(FAME, Fatty Acid Methyl Esters) is usually produced by transesteri-
fication of triglycerides, the main constituent of vegetable oil, with
methanol or ethanol. The most important parameters that affect
the rate of transesterification are reaction temperature, type and
concentration of catalyst as well as alcohol to oil molar ratio [1].
In order to improve the rate of transesterification and yield
of FAME the reaction of transesterification could be catalyzed
by homogeneous (alkalies and acids) or heterogeneous catalysts.
Nowadays, homogeneous base catalysts are the most frequently
used in industry, since the process is faster under mild reac-
tion conditions compared to acid catalyzed reaction. However,
their utilization in vegetable oils transesterification very often
forms soaps as undesirable byproducts, which in turn generates
large amounts of wastewater during the separation of the cata-
lyst and formed products. Heterogeneous catalyst could overcome
mentioned drawbacks of homogeneous catalysts; they can reduce
∗ Corresponding author. Tel.: +381 11 3303 710; fax: +381 11 3370 473.
E-mail address: skala@tmf.bg.ac.rs (D. Skala).
production costs, be reused, regenerated [2] and, finally, heteroge-
neous catalysts could be operated in continuous processes.
A variety of solid catalysts for biodiesel production has been
investigated. Those include alkaline earth base oxides [3–12], zeo-
lites and modified zeolites [13], hydrotalcites [14], and alkali or
alkaline earth oxides on porous support [15–17]. Some of them are
produced using complex and expensive procedures, which is a big
disadvantage for their industrial application.
Among the heterogeneous base catalysts, CaO is the most
studied due to its low price and desired activity (Table 1). The
catalytic activity of CaO strongly depends on calcination temper-
ature [3,5] and used precursor [3,6]. Since CaO, known to be active
in methanolysis reaction (yield over 90% after 90 min [3]), tends
in smaller extent to be leached by methanol [7] it is important to
improve its properties by fixing it to some support, e.g. silica [15],
alumina [16], or ZnO [17]. The support is usually a porous material
providing higher surface area, with catalytic activity ranging from
very small to none.
One of the catalysts showing excellent activity under moder-
ate reaction conditions (reaction time 3 h, FAME yield 94% and
catalyst can be reused up to 3 times with FAME yield above 90%
[18]) is the mixture of CaO and ZnO oxides [17–20]. It might be
obtained by calcination of calcium zinc hydroxide hydrate (calcium
zincate dihydrate – CaZn2(OH)6·2H2O) synthesized by coprecipita-
tion of ZnO and Ca(OH)2 added to 20% KOH solution [19]. Proposed
method of CaZn2(OH)6·2H2O synthesis consists of several steps.
One of them is long-lasting, as is the washing of formed calcium
0926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2012.03.032

2. ˇZ. Kesi´c et al. / Applied Catalysis A: General 427–428 (2012) 58–65 59
Table 1
Literature review of the activity of CaO, ZnO, CaO·ZnO mixed oxides and supported CaO as heterogeneous catalysts for biodiesel production.
Catalyst Oil wt% Reaction conditions Yield, % Reference
T, ◦
C
atmosphere
Molar ratio Time, h
CaO Sunflower 1 60
N2
13:1 1.5 >90 [3]
CaO Rapeseed 0.66 60 7.2:1 3 90 [8]
CaO Soybean 1 65
N2
18:1 1 90 [9]
CaO Soybean and waste cooking 1 65 14:1 2 99 [10]
CaO Soybean 1 65 12:1 2 99 [11]
CaO/SBA15 Sunflower 1 60 12:1 5 95 [15]
CaO/Al2O3 Palm 3.5 65 12:1 5 95 [16]
CaO·ZnO Sunflower 1.3 60 12:1 2 >90 [17]
CaO·ZnO
Ca/Zn ratio of 0.25
Palm kernel 10 60 30:1 1 >94 [18]
CaZn2(OH)6·2H2O Sunflower 4 60
N2
12:1 0.75 >90 [19]
CaO·ZnO Jatropha curcas 4 65 15:1 6 >80 [20]
ZnO Soybean 5 100 55:1 7 15 [21]
Table 2
Synthesized catalysts – used working condition, basicity and surface area.
Catalyst denotation Molar ratio of
Ca(OH)2 and ZnO,
medium
Preparation
method
Calcination
temperature, ◦
C
Basic strength
(H−)
Basicity
(mmol g−1
)
Surface area
(m2
g−1
)
CP Ca(OH)2:ZnO 1:2,
KOH solution
Co-precipitation / / / /
CP700 Ca(OH)2:ZnO 1:2,
KOH solution
Co-precipitation 700 9.3–10 1.31 5.7
BMH Ca(OH)2:ZnO
1:2 + H2O
Ball-millinga
7 h / / / /
BMH700 Ca(OH)2:ZnO
1:2 + H2O
Ball-milling 7 h 700 9.3–10 2.94 8.6
BM Ca(OH)2:ZnO 1:2 Ball-milling 7 h / / / /
BM700 Ca(OH)2:ZnO 1:2 Ball-milling 7 h 700 9.3–10 2.08 7.2
a
A vessel was periodically open to atmosphere while other was closed throughout milling.
zinc hydroxide hydrate with distilled water until neutral condi-
tion. Namely, the presence of KOH, even in small traces, influences
the methanolysis reaction as catalyst, which in return can make it
difficult to dissolve true catalytic activity of CaO·ZnO obtained after
calcinations CaZn2(OH)6·2H2O at appropriate temperature.
The synthesis of CaZn2(OH)6·2H2O by mechanochemical treat-
ment of Ca(OH)2 and ZnO powders in water medium with the goal
to use it as rechargeable anodic material was recently reported [22].
Mechanochemical treatment of different powders is widely used
for the activation and synthesis of broad class of materials [23,24],
including catalysts [25].
The goal of this study is to investigate synthesis of CaO·ZnO
catalyst by mechanochemical treatment of Ca(OH)2 and ZnO pow-
der mixture with the addition of required water amount, as well
as to apply the same procedure but without added water, and
subsequent calcinations at 700 ◦C. The classical coprecipitation
procedure of CaZn2(OH)6·2H2O synthesis followed by the same
calcination procedure at 700 ◦C was also performed to compare
the catalytic activity with samples obtained by mechanochemical
treatment.
2. Experimental
2.1. Catalyst preparation
Ca(OH)2 (Centrohem, Belgrade, Serbia) and ZnO (Kemika,
Zagreb, Croatia) were used for catalyst synthesis. Mechanochemical
treatment was carried out in the planetary ball mill Fritsch Pul-
verisette 5, in air atmosphere. Two zirconia vials of 500 cm3 volume
each charged with 500 g zirconia 10 mm diameter balls were used
as milling mediums. The balls to powder mass ratio was approx.
30. A powder mixture of Ca(OH)2 and ZnO, in the molar ratio of
1:2, with, as well as without, stoichiometrically required addition
of water (4.66 g) were used as starting materials for mechanochem-
ical treatment. Angular velocity of supported (basic) disc, measured
by tachometer, was 250 rpm (26.2 rad s−1).
Calcium zinc hydroxide hydrate (CaZn2(OH)6·2H2O) was pre-
pared by coprecipitation according to Ziegler and Johnson’s
procedure [26].
Prepared catalysts are denoted as follows: CP for precipitated,
BMH for ball-milled with addition of H2O and BM for ball-milled
without addition of H2O (subscript number represents the temper-
ature of calcination, i.e. 700 ◦C) (all the information related to the
preparation of BMH and BM samples is shown in Table 2).
2.2. Catalyst characterization
XRD patterns were recorded with Ital Structure APD2000 X-ray
diffractometer in Bragg–Brentano geometry using CuK␣ radiation
( = 1.5418 ˚A) and step-scan mode (range: 10–70◦ 2Â, step-time:
0.50 s, step-width: 0.1◦).
Thermogravimetric analysis (TGA) was carried out on SDT Q600
instrument in air atmosphere, with the flow rate of 100 mL min−1
at the 20 ◦C min−1 heating rate ranging from 25 to 800 ◦C.
Fourier-transform infrared (FTIR) spectra were recorded using
BOMEM (Hartmann & Braun) spectrometer. Measurements were
conducted in wave number range of 4000–400 cm−1, with 4 cm−1
resolution.

3. 60 ˇZ. Kesi´c et al. / Applied Catalysis A: General 427–428 (2012) 58–65
The sample morphology and elemental chemical analysis were
characterized at room temperature by Quanta 200 SEM system
equipped with EDS detector (FEI company, Netherlands). The accel-
erating voltage was 20 kV.
The particle size distribution was measured by laser particle size
analyzer (PSA) on Mastersizer 2000 (Malvern Instruments Ltd., UK),
which covers the particle size range of 0.02–2000 ␮m.
BET surface areas of the synthesized catalysts were carried
out according to the multipoint N2 adsorption–desorption method
using SSA-4200 Surface Area & Pore Size analyzer (Beijing Builder
Electronic Technology Co., Ltd., China). Prior to measurements, all
samples were outgassed overnight under vacuum at 573 K.
Hammett indicator experiments were conducted to determine
the basic strength of catalysts. The following Hammett indi-
cators were used: phenolphthalein (H− = 9.3), thymolphthalein
(H− = 10.0) and 4-nitroaniline (H− = 18.4). Typically, 500 mg of the
catalyst was mixed with 1 mL of Hammett indicators solution that
was diluted in 20 mL methanol. After 2 h of equilibration the color
of the catalyst was noted. The basic strength of the catalyst was
observed to be higher than the weakest indicator that under-
went the color change, and lower than the strongest indicator
that underwent no color change. To measure the basicity of solid
bases, the method of Hammett indicator-benzene carboxylic acid
(0.02 mol L−1 anhydrous ethanol solution) titration was used.
2.3. Methanolysis reaction
The catalytic activity was evaluated in the transesterificaton
of commercial edible sunflower oil (Dijamant, Zrenjanin, Serbia;
molecular weight 876.6 g mol−1, acid value of 0.202 mg NaOH g−1)
and methanol (99.5% purity, Fluka, Switzerland). All the exper-
iments were conducted in 300 cm3 batch autoclave (Autoclave
Engineers) equipped with a heater and a mixer at 60 ◦C and 1 bar,
with the molar ratio of methanol to sunflower oil of 10:1 and with
2 wt% of catalyst based on oil weight. The agitation speed was
300 rpm. The reaction samples were taken out from the reactor
at different reaction times, and after filtration 1 ␮L aliquots were
diluted with chloroform and obtained solutions analyzed by gas
chromatography (Varian 3400) with FID detector, on-column injec-
tor and fused silica capillary column (5 m × 0.53 mm film thickness
0.5 ␮m). Injector temperature was 330 ◦C, FID detector 345 ◦C, and
the temperature program of the chromatographic system was as
follows: 2 min isothermal at 50 ◦C, 50–110 ◦C with 50 ◦C min−1,
1 min isothermal at 110 ◦C, followed by 4 ◦C min−1 ramp to 170 ◦C,
1 min held at 170 ◦C, 170–340 ◦C with 20 ◦C min−1 and at the end
15 min maintained at 340 ◦C. Quantitative analysis of FAME was
done using correction factors for FAME, tri-, di- and monoglycerides
and glycerol. Calculated correction factors were used for calculating
the mass percentage of FAME.
2.4. Catalyst stability
The presence of catalyst in solution might imply to possible
homogeneous contribution to the reaction, which requires addi-
tional steps of washing and purification of the biodiesel fraction.
For this reason, an experimental procedure to evaluate lixiviation
was employed. It consisted of mixing the catalyst with methanol
under the same experimental conditions as used in the transes-
terification process but without the oil presence. After 4 h of such
treatment, the catalyst was removed by filtration, and methanol
was mixed with the necessary volume of sunflower oil, and main-
tained at 60 ◦C for 4 h. If catalysts were lixiviated, some conversion
would be observed due to the homogeneous contribution in the
system.
The solubility of the catalyst in methanol at 60 ◦C was also deter-
mined by measuring the calcium(II) and zinc(II) concentration with
Fig. 1. XRD pattern of CP (a), BM (b), BMH (c), initial Ca(OH)2 (d) and ZnO (e).
HITACHI Z-2000 polarized zeeman atomic absorption spectropho-
tometer.
The possibility to reuse the catalyst was also tested to check its
capacity to provide the same catalytic activity. The recycled use
of catalyst BMH700 was carried out. After filtration, the recovered
BMH700 catalyst was reused in the next run. Then the reaction test
was repeated 4 times using the same ratio of the oil to catalyst and
methanol to oil.
3. Results and discussion
3.1. Catalyst characterization
Fig. 1 shows the XRD patterns of coprecipitated calcium zinc
hydroxide hydrate (a), mechanochemically treated Ca(OH)2 and
ZnO with/without water addition (b and c), as well as initial
Ca(OH)2 and ZnO powders (d and e).
XRD of CP shows intense peaks of CaZn2(OH)6·2H2O (JCPDS 25-
1449) with few weak peaks assigned to ZnO (JCPDS 36-1451), con-
firming that calcium zinc hydroxide hydrate (CaZn2(OH)6·2H2O)
was synthesized (Fig. 1a).
XRD analysis of BMH reveals calcium zinc
hydroxide hydrate and zinc oxide with no detectable Ca(OH)2
(Fig. 1c). The presence of ZnO peaks in BMH indicates that the
formation of calcium zinc hydroxide hydrate was not completed.
Peak at 29.4 2Â is assigned to CaCO3 (JCPDS 5-586), which is partly
obtained by the reaction of Ca(OH)2 with CO2 from air and partly
originates from the initial Ca(OH)2 (Fig. 1d). It can be concluded
that mechanochemical reaction of Ca(OH)2 and ZnO powders
with added stoichiometrical amount of water yielded a mixture of
CaZn2(OH)6·2H2O, ZnO and CaCO3 phases.
Mechanochemical treatment of Ca(OH)2 and ZnO without the
addition of water obviously does not lead to formation of calcium
zinc hydroxide hydrate, as XRD analysis revealed (Fig. 1b). Also, it
should be noted that other calcium compounds could not be iden-
tified, probably due to amorphization of CaCO3 and Ca(OH)2 during
ball milling, showing similar mechanism to the Cu/ZnO system [27].
Thus, XRD pattern of BM sample shows only peaks of ZnO.
XRD analysis of the samples treated at 700 ◦C reveals that cal-
cination, regardless of the procedures of precursor preparation
(coprecipitation or ball milled Ca(OH)2 and ZnO powders either
with or without the addition of water) leads to the formation of
CaO (JCPDS 37-1497) and ZnO mixture, with other phase(s) being
undetectable (Fig. 2). It is very likely that fine CaO particles are

4. ˇZ. Kesi´c et al. / Applied Catalysis A: General 427–428 (2012) 58–65 61
Fig. 2. XRD pattern of CP700 (a), BM700 (b) and BMH700 (c).
embedded in ZnO matrix, suggesting that ZnO actually represents
the support for CaO.
TGA analysis of CaZn2(OH)6·2H2O is characterized by two-step
decomposition [28]. The first dominant step of weight loss could
be observed from 120 to 180 ◦C, which may be attributed to the
elimination of hydrated water and dehydration of Zn(OH)2 to form
ZnO. In respect to the initial composition of used powders, it should
be approximately 23.3%. The weight loss at about 385 ◦C corre-
sponds to the dehydratation of Ca(OH)2 (5.8%) [28]. A weight loss
at 700 ◦C (Fig. 3) indicates the presence of calcium carbonate.The
results of TG analysis for CP is close to theoretical calculation (first
weight loss up to 200 ◦C was 18.8%) and weight loss at 700 ◦C being
3.0%, corresponding to the presence of CaCO3 in the initial Ca(OH)2
(Fig. 1). In fact, TG analysis of Ca(OH)2 (not given) revealed that it
already contains some amount of CaCO3.For BMH the first weight
loss up to 200 ◦C was 12.9%, which is less than theoretical one and
in agreement with XRD results that mechanochemical synthesis
was not completed. Also, the weight loss of 7.0% at 700 ◦C indicates
that during mechanochemical treatment the reaction of CO2 and
Ca(OH)2 to CaCO3 occurred. Most likely, Ca(OH)2 reacts with CO2
from the vial environment at the very beginning of milling because
CaCO3 could not be formed by the reaction between calcium zinc
Fig. 3. TGA curves of CP, BM and BMH.
Fig. 4. FTIR spectrum of: CP (a); CP700 (b); BMH (c); BMH700 (d), BM (e) and BM700
(f).
hydroxide hydrate and CO2 from air [19]. Such conclusion was
also proven after analysis of mechanochemically prepared sam-
ples using different working procedure – one vessel was opened
every hour during 7 h of milling, thus enabling contact of ZnO and
Ca(OH)2 powder mixture with air, while other vessel being kept
closed throughout milling. The results of TG analysis revealed the
same amount of CaCO3 in both cases, indicating that formation of
CaCO3 took place at the beginning of the milling process [29]. This
is also in agreement with the result of Lopez Granados et al. [3] who
showed that the carbonation of CaO is very rapid (couple of minutes
are required to extensively carbonate the sample). Furthermore, it
is also known that CaCO3 does not catalyze the methanolysis of
triglycerides, and that its transformation into the active CaO phase
could be only realized at high temperatures.
According to TGA of BM, there were three weight-loss steps. The
first and second weight losses are the result of Ca(OH)2 decompo-
sition (about 7 wt%) and obviously above 600 ◦C decomposition of
CaCO3 occurred giving 8.0 wt% of weight loss.
FTIR spectra of the CP, CP700, BMH, BMH700 and BM catalysts are
shown in Fig. 4.
Typical bands for calcium zinc hydroxide hydrate were detected
for CP and BMH sample (Fig. 4a and c). The hydroxyl ions are char-
acterized by sharp bands appearing between 3700 and 3500 cm−1.
Therefore, two sharp bands at 3615 and 3505 cm−1, which can be
seen at Fig. 4a and c, are assigned to (OH) stretching vibrations
[30]. The weak band at 3643 cm−1 is normally ascribed to OH groups
of Ca(OH)2. The OH groups also forms a bridge between the two
metals. The bridging OH bending mode is visible at 940 cm−1 [31].
The wide band that is observed in the region of 3400–3100 cm−1 is
caused by the stretching of water molecules. Also, the HOH bending
mode of lattice water appears at 1600 cm−1, and this band exists in
all FTIR spectra, although for BM catalyst prepared without water
is very weak. The stretching bands at 3150, 3034 and 2880 cm−1
are attributed to the O H groups from H2O molecules [30]. The
band at 1070 cm−1 attributed to the Zn O H bending vibration is
observable for CP and BMH catalysts.
The presence of carbonates in all the samples is confirmed
by the broad band centered at 1465 cm−1 [19]. This band can
be assigned to stretching vibrations of O C O [3]. The bands at
874 and 712 cm−1 arises from carbonates group as well, and the
asymmetrical stretch of CO2 gives a band in the FTIR spectrum
at 2350 cm−1 [31]. The increase amount of carbonates was deter-
mined for the samples of catalysts obtained by ball-milling.

5. 62 ˇZ. Kesi´c et al. / Applied Catalysis A: General 427–428 (2012) 58–65
Fig. 5. Particle size distribution of CP700, BM700 and BMH700.
FTIR analysis of calcined samples indicates the presence of
calcium carbonate (lower amount of calcium carbonate, band at
1465 cm−1, in comparison to the amount of CaCO3 in initial Ca(OH)2
used for synthesis), as well as OH groups of Ca(OH)2 and the HOH
bending mode of lattice water (band at 3615 cm−1) which is the
result of the fact that carbonation and the hydration of CaO in air is
very rapid with just a few minutes being required to carbonate and
hydrate the sample [3]. Bands at frequencies lower than 500 cm−1
have not been explored yet, but it is supposed they refer to the
bonds in the oxides of Zn and Ca [30].
The particle size distribution of the samples prepared by calci-
nation at 700 ◦C of different precursors, namely CP700, BM700 and
BMH700 catalysts, is shown in Fig. 5.
Remarkably difference is notable between the catalyst obtained
by coprecipitation and obtained by ball milling with addition of
water. The particle size distribution of the CP700 and BM700 catalyst
is uniform with the size range of 0.2–30 ␮m, while for the BMH700
bimodal distribution is obtained: a larger fraction of the powder
particles is within the size range of 0.2–3 ␮m and the rest is within
the range of 3–40 ␮m.
The difference of median particle sizes of the CP700 and BM700
powders is obvious, while BMH700 sample has the smallest median
particle size. The median particle sizes of CP, BMH and BM samples
are 12.0, 6.2 and 3.9 ␮m, respectively (particle size distribution not
given), and after calcination at 700 ◦C decrease to 8.6 and 3.8 ␮m for
CP700 and BMH700 catalyst, respectively, while for BM700 remained
almost the same (3.8 ␮m). Such an effect could be expected because
calcination process of CP and BMH causes removal of H2O and CO2
from the CaZn2(OH)6·2H2O and CaCO3, inducing particle crush-
ing and diminution. Hence, the particle size distribution shifts to
smaller values for CP700 and BMH700 catalysts but not for BMH700
(Fig. 5).
The bimodal particle size distribution of the BMH700 catalyst,
shown in Fig. 5, can be explained by the SEM analysis (Fig. 6).
The major difference in the morphology can be seen from SEM
images of mechanochemically synthesized catalyst with water and
catalyst prepared by coprecipitation method. The SEM images of
the CP700 (Fig. 6a and b) show the existence of large, plate like par-
ticles of hexagonal shape, with particle sizes ranging from less than
1 to 20 ␮m, while BMH700 (Fig. 6e and f) shows the agglomera-
tion of small round-shape particles, including nano-particles. The
small particles were merged together giving large agglomerates.
The existence of the very small, nano sized particles along with
the large agglomerates could be the explanation for the results
obtained from particle size analysis where bimodal distribution
was determined (Fig. 5). In the BM700 sample (Fig. 6c and d) irregu-
lar plate like grains are found with small particles on them. Smooth,
polygonal plates observed at CP700 and BM700 catalysts appear to
be ZnO crystals [32] with small particles of CaO [33] dispersed on
them. It appears that better dispersion is achieved when ball milling
procedure for catalyst preparation was used, and this is the most
pronounced in the case of BMH700 where small round particles of
CaO entirely covered ZnO crystals.
Several typical points on the surface of each sample were
selected for SEM-EDS analysis to determine the atomic distribution
of Ca, Zn and O and an average value for each catalyst is considered.
The atomic Zn/Ca ratios for BM700 and BMH700 are 1.92 and 1.9,
respectively, which is close to the theoretical values corresponding
to the calcium zinc hydroxide hydrate, while determined atomic
Zn/Ca ratio for CP700 is 3.60, most likely due to the heterogeneous
distribution of ZnO and CaO [19].
Basic strength of all calcined samples, i.e. mechanochemically
obtained BMH700 and BM700 and the catalyst prepared by coprecip-
itation method CP700, was lower than H− = 18.4 and H− = 10 since
no color change was observed. All catalysts had the basic strength in
the range of 9.3–10. Higher basicity was found for BMH700 catalyst
and BM700 compared to CP700 (Table 2).
Better dispersion of CaO on the surface of ZnO could remarkably
increase the basicity of the catalyst. The increased basicity could be
attributed to the preparation method. According to the research of
Watanabe et al. the surface basicity of the alkaline earth hydroxides
could be increased by milling [34]. Obviously, basicity of BMH700
is the highest, and it could influence the catalytic activity for the
biodiesel synthesis. Namely, the reaction activity depends on the
number of basic sites present in the catalysts as well as on their
strength [35].
BET analysis was carried out in order to determine the specific
surface area of the catalysts used in this study (Table 2). The ball
milled and calcined catalysts exhibited a slightly higher surface
area than the CP700. However, the difference of specific surface
area between prepared samples of catalyst is not crucial for their
catalytic activity indicating that all prepared samples of CaO·ZnO
can be characterized as catalysts with small surface area and low
porosity.
3.2. Activity of synthesized catalysts
Beforehand, the catalytic behavior of non-calcined samples
was also investigated. CP catalyst dried at 100 ◦C for 45 min (to
remove H2O absorbed in contact with air) did not show activity
in methanolysis of sunflower oil. Drying procedure is necessary
since the presence of water is not desirable because it influ-
ences the saponification reaction. The removal of water from wet
coprecipitated CaZn2(OH)6·2H2O suggests that crystal water and
corresponding hydroxides could not catalyze reaction between
methanol and triglycerides.
BMH also showed poor activity (23% yield of FAME after 4 h),
but it was still noticeable better comparing to precipitated cal-
cium zinc hydroxide hydrate. Martín Alonso et al. [6] reported that
neither CaCO3 nor Ca(OH)2 were active in transesterification of
triglycerides with methanol, while Kouzu et al. [11] pointed out
that Ca(OH)2 showed some but very low activity (12% yield after
1 h of reaction).
The initial idea to obtain CaO·ZnO mixture from ball milled
Ca(OH)2 and ZnO powders without the addition of water (BM)
which poses desirable catalytic activity was also tested in this study.
However, the sample BM did not show activity in methanolysis
reaction, indicating that Ca(OH)2 was not transformed to active CaO
during milling. Such finding is in agreement with the XRD result,

6. ˇZ. Kesi´c et al. / Applied Catalysis A: General 427–428 (2012) 58–65 63
Fig. 6. SEM images of CP700 (a) and (b); BM700 (c) and (d); and BMH700 (e) and (f).
which shows that only ZnO phase was identified on XRD pattern
(Fig. 1).
Finally, the FTIR analysis proved that crystal water, hydroxides,
as well as formed carbonates could be removed or transformed into
corresponding oxides only after calcination at 700 ◦C (activation of
BMH, BM and CP).
It is clear from activity test of prepared catalysts for biodiesel
synthesis presented in Fig. 7 that all samples were active for trans-
esterification reaction of sunflower oil with methanol after their
activation at 700 ◦C. The calcination step at 700 ◦C is necessary for
obtaining desired activity of BMH, BM and CP samples.
The catalysts prepared by different methods showed different
activity. The best conversion of triglycerides to fatty acid methyl
esters with 92% yield of FAME after 2 h and over 97.5% after 4 h was
achieved with BMH700 catalyst. The high yield was also obtained
with BM700 catalyst, but longer time was required (3 h) to reach
over 90% the FAME content in an oil phase. The obtained results
clearly indicate that BMH700 catalyst is more active, and that
shorter time is necessary to reach desired composition of reac-
tion mixture compared to BM700 catalyst. The catalyst prepared
by coprecipitation method was less active giving only 84% of FAME
in oil phase after 4 h. The prepared sample BMH was also calcined
at 400 ◦C and BMH400 tested in the methanolysis of sunflower oil;
after 4 h of reaction, the obtained yield of FAME was 89.6%. Activ-
ity of BMH400 and obtained FAME yield was used to determine
the appropriate temperature for calcination of BMH sample being
700 ◦C (BMH700 gave 97.1% FAME yield after 3 h). Such decision was
proven by TG analysis of BMH sample which indicates the presence
of larger amount of calcium carbonate which is the reason why
calcination at higher temperature is necessary.
Although the XRD analysis of CP, BMH and BM catalysts has
shown that for CP and BMH procedure of catalyst synthesis, the
CaZn2(OH)6·2H2O was synthesized, which is not in the case for
BM catalyst preparation. In all cases CaO·ZnO powder mixture was
obtained after calcination at 700 ◦C. Thus, the difference in the
activity of both mechanochemically synthesized catalysts and cat-
alyst prepared by coprecipitation procedure could be related to the
different particle size and particle size distribution, as well as to
the alkalinity of prepared catalysts. The results of biodiesel synthe-
sis confirmed such assumption that simple relation between the

7. 64 ˇZ. Kesi´c et al. / Applied Catalysis A: General 427–428 (2012) 58–65
Fig. 7. FAME yields for catalysts synthesized by coprecipitation or ball milling and
calcined at 700 ◦
C; experimental conditions: 60 ◦
C and 1 bar, molar ratio of methanol
to sunflower oil of 10:1 and with 2 wt% of catalyst.
particles size and time necessary to reach the same conversion of
triglycerides exists. Obviously, the rate of reaction increased with
decreasing particle sizes, as the effect of larger external surface
area of catalyst and decreased resistance for triglyceride diffusion
from the bulk liquid phase to that surface. Taking the alkalinity
and the catalytic activity into account, it could be concluded that
higher alkalinity of prepared catalyst increases the conversion of
triglyceride.
3.3. Catalyst stability
A key aspect in the development of solid catalysts for heteroge-
neous biodiesel synthesis is avoiding undesired lixiviation of active
species. If lixiviation is high, the active species could also act as a
homogeneous catalyst and thus, the advantages of the heteroge-
neous catalyst would not really exist. The previous studies have
demonstrated that bulk CaO is partially dissolved in the methanol
[7]. To evaluate lixiviation, catalyst BMH700 was placed in contact
with methanol under the same experimental conditions as used
in the transesterification process of sunflower oil (triglycerides).
The catalyst was removed by filtration and then the methanol was
placed in the contact with the vegetable oil, and maintained at 60 ◦C
for 4 h. The analysis of the resulting solution revealed a FAME yield
of 0.51%, indicating that lixiviation was negligible. The FAME yield
of 0.51% in 4 h can be associated to the very small amount of solid
catalyst, which is not removed from methanol after filtration.
The data of catalyst solubility in methanol is very important for
further application of solid catalyst. The reaction will be homoge-
neous if the catalyst is soluble in reactants. The amount of Ca2+
ion that can be present in methanol has been investigated after
2 h catalysts with methanol at 60 ◦C contact and catalysts removal.
The result showed that the concentrations of dissolved Ca2+ from
CP700, BM700 and BMH700 catalysts were 4.32, 48.7 and 8.90 mg L−1,
respectively. López Granados et al. [7] reported that the solubility
of CaO was 134 and 169 mg L−1 CaO in methanol after 1 and 3 h
contact time, respectively. Expressed as Ca2+ concentration, it is
equivalent to 96 and 121 mg L−1of Ca2+ methanol. It indicates that
all the catalysts sample prepared in this study are more stable than
pure CaO, thus suggesting that mixed CaO and ZnO oxides stabilized
the CaO or active phase and decrease the rate of leaching process.
To found out whether deactivation of the active sites due to their
poisoning by adsorption of triglycerides or some other molecule
present in the reaction mixture occurs, used catalyst BMH700 was
filtered and reused in several repeated biodiesel synthesis under
the same reaction conditions. Achieved FAME yield after fourth run
was 90%. Therefore, it was concluded that BMH700 have good activ-
ity and might be reused several times. The reason for stability of
CaO could be attributed to its strong interaction with less soluble
ZnO.
4. Conclusion
The synthesis of CaO·ZnO mixtures was carried out and tested in
reaction of methanol and sunflower oil. Different samples of CaO
precursor were obtained by: (a) ball milling of Ca(OH)2 and ZnO
powders with or without the addition of water, and (b) coprecip-
itation of ZnO and Ca(OH)2 added to 20% KOH solution. Prepared
precursors were calcinated at 700 ◦C leading to the formation of
CaO·ZnO mixtures in all cases. After the calcination at 700 ◦C, cat-
alyst obtained by ball milling was more active compared to the
catalyst calcined prepared by coprecipitation at the same temper-
ature. The reasons for the different activity of synthesized catalysts
could be explained by the difference of their basicity and the differ-
ence of their particle sizes. These facts imply the importance of the
preparation procedure, which significantly affect the properties of
the catalysts, as well as the catalytic activity.
The highest catalytic activity exhibits the catalyst obtained by
mechanochemical treatment of Ca(OH)2 and ZnO powders with
added water, and subsequent calcination at 700 ◦C. Such prepared
catalyst gave FAME formation of 97.5% after 4 h reaction of sun-
flower oil and methanol (1:10 molar ratio) in a batch reactor at
60 ◦C. Also, the results showed that the CaO·ZnO catalysts were
practically insoluble in methanol indicating that CaO properties are
improved by mixing with ZnO. Repeated test of biodiesel synthesis
with same amount of BMH700 catalyst showed that catalyst keep
acceptable activity. Namely, decrease of FAME yield from 97.1%
after first to 90% after fourth use of same amount of catalyst was
experimentally determined.
Acknowledgments
This work was financially supported by the Ministry of Edu-
cation and Science of the Republic of Serbia (Grant No. 45001).
The authors thank M.Sc. Zoran Stojanovi´c ITS SASA for performing
particle size measurements and to Prof. Nevenka Raji´c for valu-
able comments related to TG/DTG analysis and characterization of
prepared samples of catalysts.
References
[1] Y.C. Sharma, B. Singh, S.N. Upadhyay, Fuel 87 (2008) 2355–2377.
[2] S. Gliˇsi´c, I. Luki´c, D. Skala, Bioresour. Technol. 100 (2009) 6347–6354.
[3] M. Lopez Granados, M.D. Zafra Poves, D. Martın Alonso, R. Mariscal, F. Cabello
Galisteo, R. Moreno-Tost, J. Santamaría, J.L.G. Fierro, Appl. Catal. B 73 (2007)
327–336.
[4] M. Di Serio, M. Ledda, M. Cozzolino, G. Minutillo, R. Tesser, E. Santacesaria, Ind.
Eng. Chem. Res. 45 (2006) 3009–3014.
[5] V. Veljkovic, O. Stamenkovic, Z. Tododrovic, M. Lazic, D. Skala, Fuel 88 (2009)
1554–1562.
[6] D. Martín-Alonso, F. Vila, M. Ojeda, M. López Granados, J. Santamaría-González,
Catal. Today 158 (2010) 114–120.
[7] M. López Granados, D. Martín-Alonso, I. Sadaba, R. Mariscal, P. Ocon, Appl. Catal.
B 89 (2009) 265–272.
[8] A. Kawashima, K. Matsubara, K. Honda, Bioresour. Technol. 100 (2009)
696–700.
[9] M. Kouzu, T. Kasuno, M. Tajika, S. Yamanaka, J. Hidaka, Appl. Catal. A 334 (2008)
357–365.
[10] M. Kouzu, S. Yamanaka, J. Hidaka, M. Tsunomori, Appl. Catal. A 355 (2009)
94–99.
[11] M. Kouzu, T. Kasuno, M. Tajika, Y. Sugimoto, S. Yamanaka, J. Hidaka, Fuel 87
(2008) 2798–2806.
[12] G. Arzamendi, E. Arguinarena, I. Campo, S. Zabala, L.M. Gandia, Catal. Today
133–135 (2008) 305–313.

8. ˇZ. Kesi´c et al. / Applied Catalysis A: General 427–428 (2012) 58–65 65
[13] G.J. Suppes, M.A. Dasari, E.J. Doskocil, P.J. Mankidy, M.J. Goff, Appl. Catal. A 257
(2004) 213–223.
[14] W. Xie, H. Peng, L. Chen, J. Mol. Catal. A: Chem. 246 (2006) 24–32.
[15] M.C.G. Albuquerque, I. Jiménez-Urbistondo, J. Santamaría-Gonzáles, J.M.
Mérida-Robles, R. Moreno-Tost, E. Rodríguez-Castellón, A. Himénez-López,
D.C.S. Azevedo, C.L. Cavalcante Jr., P. Maireles-Torres, Appl. Catal. A 334 (2008)
35–43.
[16] M. Zabeti, W.M.A.W. Daud, M.K. Aroua, Fuel Process. Technol. 91 (2010)
243–248.
[17] A.C. Alba-Rubio, J. Santamaría-Gonzáles, J.M. Mérida-Robles, R. Moreno-Tost, D.
Martín-Alonso, A. Jiménez-López, P. Maireles-Torres, Catal. Today 149 (2010)
281–287.
[18] C. Ngamcharussrivichai, P. Totarat, K. Bunyakiat, Appl. Catal. A 341 (2008)
77–85.
[19] J.M. Rubio-Caballero, J. Santamaría-Gonzáles, J.M. Mérida-Robles, R. Moreno-
Tost, A. Jiménez-López, P. Maireles-Torres, Appl. Catal. B 91 (2009) 339–346.
[20] Y.H. Taufiq-Yap, H.V. Lee, M.Z. Hussein, R. Yunus, Biomass Bioenergy 35 (2011)
827–834.
[21] W.M. Antunes, C.O. Veloso, C.A. Henriques, Catal. Today 133–135 (2008)
548–554.
[22] X.-M. Zhu, H.-X. Yang, X.-P. Ai, J.-X. Yu, Y.-L. Cao, J. Appl. Electrochem. 33 (2003)
607–612.
[23] E. Avvakumov, M. Senna, N. Kosova, Soft Mechanochemical Synthesis, Kluwer
Academic Publishers, 2001.
[24] P. Baláˇz, Mechanochemistry in Nanoscience and Minerals Engineering,
Springer-Verlag, Berlin, Heidelberg, 2008.
[25] S. Petrovi´c, A. Terlecki-Bariˇcevi´c, Lj. Karanovi´c, P. Kirilov-Stefanovi´c, M.
Zduji´c, V. Dondur, D. Paneva, I. Mitov, V. Raki´c, Appl. Catal. B 79 (2008)
186–198.
[26] F. Ziegler, C.A. Johnson, Cem. Concr. Res. 31 (2001) 1327–1333.
[27] H.L. Castricum, H. Bakker, B. van der Linden, E.K. Poels, J. Phys. Chem. B 105
(2001) 7928–7937.
[28] S. Wang, Z. Yang, L. Zeng, Mater. Chem. Phys. 112 (2008) 603–606.
[29] A.M. Kalinkin, E.V. Kalinkina, O.A. Zalkind, T.I. Makarova, Inorg. Mater. 41 (2005)
1073–1079.
[30] T. Lin, M.A. Mollah, R. Vempati, D. Cocke, Chem. Mater. 7 (1996) 1974–1978.
[31] K. Nakamoto, Infrared, Raman Spectra of Inorganic and Coordination Com-
pounds, 5th ed., John Wiley & Sons, New York, 1997.
[32] F. Liu, Y. Zhang, Ceram. Int. 37 (2011) 3193–3202.
[33] Dj. Vujicic, D. Comic, A. Zarubica, R. Micic, G. Boskovic, Fuel 89 (2010)
2054–2061.
[34] T. Watanabe, J. Liao, M. Senna, J. Solid State Chem. 115 (1995) 390–394.
[35] S. Bancquart, C. Vanhove, Y. Pouilloux, J. Barrault, Appl. Catal. A 218 (2001)
1–11.