Transesterification of triacetin with methanol on various solid acid catalysts: A role of catalyst properties

Article abstract of DOI:10.1016/j.apcata.2010.07.060

Present studies concentrated on a comparison of catalytic performance of number of solid acid catalysts, namely Amberlyst-15, Nafion-SAC-13, polyaniline-sulfate, silver and cesium salts of H₃PW ₁₂O₄₀ in transesterifiaction of triglycerides with methanol. Reaction was studied for triacetin, the short triglyceride and for castor oil consisting of vegetable fatty acid esters under mild reaction conditions (atmospheric pressure, 50 °C). The catalysts were characterized by measuring textural (specific surface area, porosity, pore diameter) and acid (concentration and the strength of acid centers) properties. The affinity of catalyst to highly hydrophobic triglyceride reactant relative to that toward rather hydrophilic glycerol was also evaluated from the measurements of contact angles for droplets of triacetin, castor oil and glycerol on the surface of pressed tablet of catalyst. This affinity of catalysts was also evaluated from the drop of catalyst activity induced by glycerol added to the reaction mixture or by pre-treatment of catalyst with triacetin. Cs₂HPW ₁₂O₄₀ salt was the most active catalyst, but it deactivated in the transesterification. The highest strength of acid sites and high affinity to glycerol was determined for this Cs salt. Thus, affinity of studied catalysts toward glycerol and the strength of acid sites seem to be decisive factors for their activity and in particular for deactivation due to polar reagents adsorption. Greater affinity toward glycerol and higher strength of acid sites, deactivation of catalysts would be easier.

Full text of DOI:10.1016/j.apcata.2010.07.060

Applied Catalysis A: General 387 (2010) 13–25
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Transesterification of triacetin with methanol on various solid acid catalysts:
A role of catalyst properties
a
a,∗
b
a
c
A. Zi e˛ ba , A. Drelinkiewicz , P. Chmielarz , L. Matachowski , J. Stejskal
a
Institute of Catalysis and Surface Chemistry, Polish Academy of Sciences, Niezapominajek 8, 30-239 Kraków, Poland
Faculty of Chemistry, Technical University, 35-059 Rzeszów, W. Pola 2, Poland
Institute of Macromolecular Chemistry, Academy of Sciences of the Czech Republic, Heyrovsky Sq. 2, 162 06 Prague 6, Czech Republic
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 28 May 2010
Received in revised form 28 July 2010
Accepted 28 July 2010
Available online 6 August 2010
Present studies concentrated on a comparison of catalytic performance of number of solid acid catalysts,
namely Amberlyst-15, Nafion-SAC-13, polyaniline-sulfate, silver and cesium salts of H3PW12O40 in trans-
esterifiaction of triglycerides with methanol. Reaction was studied for triacetin, the short triglyceride and
for castor oil consisting of vegetable fatty acid esters under mild reaction conditions (atmospheric pres-
◦
sure, 50 C). The catalysts were characterized by measuring textural (specific surface area, porosity, pore
diameter) and acid (concentration and the strength of acid centers) properties. The affinity of catalyst
to highly hydrophobic triglyceride reactant relative to that toward rather hydrophilic glycerol was also
evaluated from the measurements of contact angles for droplets of triacetin, castor oil and glycerol on the
surface of pressed tablet of catalyst. This affinity of catalysts was also evaluated from the drop of catalyst
activity induced by glycerol added to the reaction mixture or by pre-treatment of catalyst with triacetin.
Cs2HPW12O40 salt was the most active catalyst, but it deactivated in the transesterification. The highest
strength of acid sites and high affinity to glycerol was determined for this Cs salt. Thus, affinity of studied
catalysts toward glycerol and the strength of acid sites seem to be decisive factors for their activity and in
particular for deactivation due to polar reagents adsorption. Greater affinity toward glycerol and higher
strength of acid sites, deactivation of catalysts would be easier.
Keywords:
Transesterification
Bio-esters
Triacetin
Castor oil
Acid catalysts
©
2010 Elsevier B.V. All rights reserved.
1
. Introduction
free fatty acid and transesterification of triglycerides. Their catalytic
efficiency has been reviewed in recently published papers [1–5].
Among heterogeneous acid catalysts the samples based on inor-
ganic oxides namely sulfated zirconia [6–9], tungstated zirconia
[8,10,11], strong mineral acids such as Keggin type heteropoly-
Biodiesel fuel, an alternative to classic diesel fuel, belongs to
ecological fuels because it consists of methyl esters of fatty acids,
derived from vegetable oils or animal fats. Methyl esters are
obtained by the transesterification of triglycerides with short-chain
alcohols, such as methanol. The reaction is catalysed by strong acids
or bases, but owing to environmental, technological, and economic
reasons, the substitution of industrial homogeneous catalysts by
heterogeneous catalysts is a desirable goal.
The advantage of solid catalysts consists in the ease of prod-
uct separation and recycling of the catalyst. Furthermore, solid
catalysts can be used in a packed bed continuous flow reactor, sim-
plifying the product separation and purification. Moreover, solid
acid catalysts are preferred because they are able to catalyse both
transesterification of triglycerides and esterification of free fatty
acids present in oil feedstock. A variety of solid acids has already
been tested for methyl ester synthesis via both, esterification of
acid H PW12O40 supported by zirconia, silica, alumina, activated
3
carbon [12] and Ta O5 [13] and its cesium salt Cs2.5H0.5PW12O40
2
[14] were mainly studied. Recently, a new class of sulfonated car-
bons catalysts, termed as “sugar catalysts” was reported. They
were prepared by incomplete carbonization of simple carbohy-
drates (starch, cellulose, glucose, sucrose) followed by sulfonation
[15,16]. Sulfonic acids – based ion-exchange resins are another
class of solid acid catalysts known to be effective in the for-
mation of methyl esters via esterification and transesterification
reactions. They are styrene-based sulfonic acid, Amberlyst-series,
and perfluorosulfonic acid-based resins, Nafion and Nafion SAC-13
composite [5,17,18], (polystyrene-divinylbenzene)-based sulfonic
acid [19], sulfonated poly(styrene-co-divinylbenzene) and sul-
fonated poly(divinylbenzene) [20].
Some common problems encountered for solid acid catalysts
concern appropriate morphological, textural and acid character-
istic of the samples. Apart from them, hydrophilic character of
catalyst surfaces and active site leaching resulting in deactivation
∗ Corresponding author.
E-mail addresses: ncdrelin@cyf-kr.edu.pl, drelinki@chemia.uj.edu.pl
A. Drelinkiewicz).
(
0
926-860X/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2010.07.060

1
4
A. Zi e˛ ba et al. / Applied Catalysis A: General 387 (2010) 13–25
Scheme 1. Transesterification of triacetin with methanol.
of catalysts are frequently discussed. Other reason of catalyst deac-
tivation is partial blockage of acid centers due to accumulation
of polar reagents like glycerol and/or partial glycerides (di- and
mono-) inside pore structure of catalysts [5,21]. Thus, to avoid polar
reagents accumulation, appropriate textural/morphological char-
acteristics and hydrophobic nature of catalyst surface to promote
the preferential adsorption of triglycerides are needed. This was
achieved by incorporation of hydrophobic organic groups (alkyl
and phenyl) into mesoporous silica materials. More hydrophobic
environment around active centres was created in which water
molecules, product of fatty acids esterification were easier excluded
from the mesopores enhancing the activity of catalysts [21]. Sim-
ilar strategy of tuning hydrophobic/hydrophilic balance of the
catalysts was also successfully adopted in case of catalysts for
transesterification of triglycerides [22,23]. It was demonstrated
that modification of support with alkyl or phenyl groups in hybrid
catalyst Ta O5/SiO –[H PW12O40/R] R = Me, Ph, resulted in higher
alyst. This effect is studied and discussed in the present work
using various solid acid catalysts. Polymeric catalysts Amberlyst-
15, Nafion-SAC-13, polyaniline-sulfate (PANI-S) and cesium and
silver salts of Keggin type heteropolyacid, H PW12O40 (HPW),
3
namely Cs HPW12O40 (abbreviated as Cs-2), Cs2.5H0.5 PW12O40
2
(abbreviated as Cs-2.5) and Ag HPW12O40 (abbreviated as Ag-2)
2
were tested.
Commercial resins, Amberlyst-15 and Nafion-SAC-13 are well
known to be effective solid acid catalysts in esterification of fatty
acids and transesterification of triglycerides. Polyaniline salts, a
hydrochloride, sulfate, nitrate, p-toluenesulfonate and phosphate,
have been studied by Palaniappan et al. [25,26] as the solid acid
catalysts for various organic transformations, among them for
the esterification of carboxylic acids with alcohols. Polyaniline-
sulfate (PANI-S) was also studied for the transesterification of
triglycerides in our previous work [27]. The activity of PANI-S was
found to be almost stable during recycling catalytic tests in ester-
ification as well as transesterification reactions [26,27]. Among
2
2
3
activity and much lower deactivation for simultaneous ester-
ification and transesterification of soybean oil with methanol
CsxH3 PW12O40 salts of various Cs content ranging from x = 1 up to
−x
compared to alkyl-free HPW/Ta O5 catalyst [13,22]. Similar effect
x = 3, the Cs2.5H0.5PW12O40 salt is known to be the most active acid
catalyst for various reactions, performed in gas-phase like isomer-
ization, alkylation, dehydratation [28,29]. This excellent activity of
Cs2.5H0.5PW12O40 was related to its superacidity, high surface area
and mesoporous structure [30]. This salt was also most active than
2
was obtained by incorporation of organosulfonic groups into
mesostructured silica SBA-15 [23,24]. In esterification of free fatty
acids present in beef tallow the propylsulfonic acid-functionalized
SBA-15 catalyst deactivated due to the condensation of reagents in
the mesopores of catalysts. This was explained by too high polar-
ity of the catalyst. Bifunctionalization of the SBA-silica support via
simultaneous incorporation of propylsulfonic groups and propyl
groups resulted in enhanced hydrophobicity of the catalyst thereby
diminishing the interaction of the polar molecules with the sulfonic
acid groups, a reason of catalyst deactivation. Very promising per-
formance of propylsulfonic acid-modified SBA-15 catalyst towards
simultaneous esterification of fatty acids and transesterification of
triglycerides using refined and crude vegetable oils was demon-
strated by Melero et al. [24].
2−
H-ZSM-5, SiO –Al O , SO
/ZrO , Nb O5 in liquid phase reactions,
2
2
3
4
2
2
such as hydrolysis of ethyl acetate and 2-methylphenyl acetate,
esterification of acrylic acid with n-butanol [30]. Cs HPW12O40
2
salt exhibits also strong acid centres, however due to very low
2
surface area (ca. 1 m /g) and microporous structure its activity
in gas phase reaction was very low, distinctly lower than that of
Cs2.5H0.5PW12O40. Both such salts were shown to be active catalysts
for transesterification of triglycerides [31,32] however, another
order of activity was observed. In methanolysis of tributyrin and
esterification of palmitic acid Narasimharao et al. [30] observed
that Cs salts with Cs content x = 2–2.3 were much more active than
Cs-2.5. Analogous relation was also obtained in our previous studies
[32]. The activity of Cs-2 was higher than that of Cs-2.5 in methanol-
ysis of triacetin and castor oil. This relation was explained taking
into account that the salt of HPW, such as Cs, Ag and K formed
colloidal dispersion in reaction medium during transesterification
of triglycerides, mostly due to the presence of methanol. Due to
In view of these literature observations, relative affinity of
catalyst to non-polar triglyceride and more polar intermediates
(
diglyceride, monoglyceride) and products, glycerol is an impor-
tant property. Weak adsorption of glycerol molecules and strong
adsorption of triglyceride molecules are preferred. However, it
can be expected that these adsorption properties would be to
some extent determined by the strength of acid sites in the cat-

A. Zi e˛ ba et al. / Applied Catalysis A: General 387 (2010) 13–25
15
open pore structure of colloidal particles, the role of specific sur-
face area and porosity determined for solid-powdered samples was
less important. Colloidal particles facilitated the access of reagents
to protons of the salts and activity was related to the bulk (formal)
acidity of the Cs salts similarly to the relation reported by Izumi
et al. [33] for the hydrolysis of ethyl acetate under excess of water.
Under these conditions their activity decreased monotonously with
increasing Cs content or decreasing acidity. In our previous work,
silver salts, namely Ag HPW12O40 and Ag2.5H0.5PW12O40 were also
2
tested for methanolysis of triglycerides [34]. They also formed
colloidal particles in reaction mixture and Ag-2 salt was more
active than Ag-2.5. The catalysts Amberlyst-15, Nafion-SAC-13,
polyaniline-sulfate, Ag-2, Cs-2 and Cs-2.5 tested in the present
work displayed good stability under methanolysis reactions owing
to high resistant for active sites leaching [26,27,31].
Here, methanolysis was studied for triacetin (glycerol triac-
etate), a model triglyceride molecule and for vegetable oil, castor
oil. The methanolysis of triacetin (the acetic acid triester of glyc-
erol) which is the simplest triglyceride has already been studied
as a model reaction for the transesterification of natural oils
[
18,35–39]. Triacetin, having the same chemical functionality of
any triglyceride molecule, shares the same reactivity principles
of triglycerides and its methanolysis is accomplished via three
sequential steps, similarly to the methanolysis of other triglyc-
erides. Reaction pattern reported by Lopez et al. [35] (Scheme 1)
shown that methanolysis proceeds via the intermediate prod-
ucts, diacetin and monoacetin to produce finally methyl acetate
and glycerol. Castor oil derived from Riccinus communis plants is
mentioned frequently in the literature as a potential raw mate-
rial for biodiesel and its transesterification was studied by Plentz
Meneghetti et al. [40,41]. The main constituent of castor oil is
triglyceride of 12-hydroxy-9-octadecenoic acid (ricinoleic acid).
Due to the presence of OH group at C-12 carbon, the castor oil
exhibits good solubility in methanol and it applies also to methyl
esters formed after its transesterification. Similarly, triacetin is also
readily soluble in methanol and methyl esters formed. Thus, during
methanolysis of both our triglycerides, there was a single liquid-
phase system. Moreover, by using a triacetin, the model compound,
we are able to gain some insight into the role of catalyst properties
in the course of consecutive reactions yielding partial glycerides,
diacetin and monoacetin.
It should be pointed out that all catalysts selected for the
present work have been previously studied for the transesterifi-
cation of triglycerides. Literature data and in particular the results
of Dr. Goodwin group proved no leaching of active centers in case
of polymeric catalysts, Nafion-SAC-13 and Amberlyst-15 [18,35].
In recycling methanolysis experiments they slowly deactivated,
however due to partial blockage of active sites by reagents. The
heterogeneous mode of HPW salts was also proved by results of
Narasimharao et al. [31] and Chai [14]. The stability of polyaniline-
sulfate catalysts was demonstrated by Palaniappan et al. [25,26]
in various esterification reactions and in our earlier work in the
transesterification of triacetin and castor oil [27].
Fig. 1. Pores distribution diagrams for polymeric catalysts PANI-S, Amberlyst-15,
Nafion-SAC-13.
2
. Experimental
The cesium and silver salts of H PW12O40, namely Cs HPW12O40
3
2
2.1. Catalysts preparation
(Cs-2), Cs2.5H0.5PW12O40 (Cs-2.5) and Ag HPW12O40 (Ag-2) were
2
prepared by means of commonly used procedure [42]. The stoichio-
metric quantities of cesium carbonate, or silver nitrate were added
to aqueous solutions of the 12-tungstophosphoric acid (Aldrich).
The concentrations of HPW and cesium/silver salt solutions were
0.10 and 0.04 M, respectively. The Cs or Ag content (x) in final salts
Amberlyst-15 and Nafion-SAC-13 were commercial reagents
Aldrich). Polyaniline sulfate (PANI-S) powder was prepared by the
(
oxidation of 0.1 M aniline sulfate with 0.25 ammonium peroxy-
disulfate in 50 vol.% ethanol–water at room temperature. The solids
were collected after 24 h on filter, rinsed with 0.1 M sulfuric acid,
acetone, and dried. The details of preparation were described in our
previous paper [27].
was adjusted by the amount of precipitating solutions (Cs CO3 or
2
AgNO ) added. The preparation was performed at ambient tem-
3
perature without stirring. The resulting white colloidal solution of

1
6
A. Zi e˛ ba et al. / Applied Catalysis A: General 387 (2010) 13–25
◦
Cs, Ag salts was slowly evaporated overnight in the oven at 40 C to
In the course of catalytic tests the samples of reaction mix-
ture were periodically taken and analysed by GC method following
the method described earlier [32]. From the data of GC analysis
the triacetin conversion (CTG), and the yields of partial glycerides,
◦
dryness. Prior to the catalytic tests the samples were dried at 120 C
in order to remove water of crystallization.
2.2. Characterization of catalysts
diacetin, monoacetin (Y ) were calculated as follows:
i
The specific surface areas of samples were calculated from the
nitrogen adsorption–desorption isotherms at 77 K in an Autosorb-
NTG,0 − NTG
CTG
=
× 100%
(1)
(2)
N
TG,0
1
, Quantachrome equipment. Prior to the measurements, the
◦
samples were preheated and degassed, under vacuum at 60 C for
Ni
Yi =
× 100%
1
8 h. Morphology of polyaniline–sulfate was studied by means of
NTG,0
Field Emission Scanning Electron Microscope JEOL JSM-7500 F.
The total amount of Brönsted acid sites was determined by acid-
base titration following procedure previously described [18,27].
An amount of ca. 0.1 g of each sample was suspended in 20 cm3
of NaOH (0.1 M) for 24 h at room temperature. The liquid sample
obtained after polymer filtration was subsequently titrated with
where NTG,0 and NTG are the moles of triacetin initially present in
the reactor and the moles of triacetin remaining at time t, respec-
tively; NDI and NMONO are the numbers of moles of diacetin and
monoacetin in the reactor at time t. The number of glycerol moles
formed (NGL) was calculated from the mass balance:
0.1 M HCl.
Acid sites strength was determined with Hammett indicators
NGL = NTG,0 − (NTG + NDI + NMONO
)
(3)
by means of procedure used by other authors [43,44]. Approx-
imately 25 mg of sample was shaken with 1 cm3 of a solution
of Hammett indicators diluted in methanol and left to equili-
brate for 2 h after which no further color changes were observed.
The color of the sample was then noted. The following Hammett
indicators were used: 4-(dimethylamino)azobenzene (pKa = +3.3),
The selectivity to diacetin (SDACT), monoacetin (SMONO) and to
glycerol (SGL) was calculated from the formula:
Ni
Si =
× 100%
(4)
NTG,0 − NTG
4
-phenylazodiphenylamine (pKa = +1.5), crystal violet (pKa = + 0.8),
dicinnamalacetone (pKa = −3), anthraquinone (pKa = −8.2), 4-
nitrotoluene (pKa = −11.35), 3-nitrotoluene (pKa = −11.99), 1-
chloro-4-nitrobenzene (pKa = −12.70), 1-chloro-3-nitrobenzene
As a measure of catalyst activity initial rate of triacetin conversion
calculated at conversion below 10%) related to 1 g of catalyst was
assumed [r, mol TACT/min g].
(
(
pKa = −13.16), 1.4-dinitrotoluene (pKa = −13.75). The acid strength
The transesterification of castor oil with methanol was performed
is quoted as being stronger than the weakest indicator which
exhibits a color change, but weaker than the strongest indicator
that gives no change.
◦
3
at 60 C using 6 g of castor oil, 7.6 cm of methanol (MR = 29) and
0
.6 g of catalyst. This amount of catalyst corresponds to catalysts
3
concentration in the reaction mixture equal to 43 g/dm .
The contact angles (␣) were measured with use of an opti-
cal goniometer (Cobrabid Optica, Poland) with a digital camera
installed in axial extension of its lens. The liquid droplets of the con-
stant volume (3–5 ␮l) were introduced onto the surface of studied
samples with micropipette. The catalyst was in the form of pressed
tablet, ca. 0.5 cm in diameter and 0.1–0.2 cm thick. The measure-
Transesterification of castor oil at the same conditions as in
◦
3
case of triacetin (50 C, low concentration of catalyst, 15.8 g/dm )
resulted in very low yield of methyl esters, practically not measur-
able for less active Cs-2.5 salt. This is a consequence of definitively
lower rate of reaction involving bulky triglycerides of castor oil.
Therefore, in order to have similar yields of methyl esters, ca.
◦
ment was taken at 21 C. The values of contact angles were found
4
0–60% after 3 h of methanolysis of both triglycerides, the trans-
from the geometric analysis of pictures taken for liquid droplets,
which involved the use of original software developed by Kontrast
◦
esterification of castor oil was carried out at temperature of 60 C
using higher concentration of catalysts, 43 g/dm . This concentra-
3
(
Pasł e˛ k, Poland) for the interpretation of Young’s equation. Nine (9)
tion was chosen in experiments with the most active Ag-2 catalyst.
The progress of castor oil methanolysis was monitored by GC
analysis of methyl esters formed [32,34]. Castor oil was entirely
composed of a triglyceride of ricinoleic acid (87.44 wt. %), traces
of glycerol (below 0.1 wt %), di- and mono-glycerides of rici-
noleic acid, and free ricinoleic acid (below 0.1 wt. %). From the
GC analysis of methyl esters apart from the dominating triglyc-
erides of ricinoleic acid, low amounts of triglycerides of other fatty
acids (linoleic 5.05 wt.%, oleic 3.88 wt.%, stearic 1.4 wt.%, palmitic
droplets were analysed each time which were placed on the surface
simultaneously. For each picture recorded, the geometrical shape
analysis was repeated ten times.
2.3. Catalytic tests
The transesterification of triglycerides, triacetin (Fluka) and cas-
tor oil (Microfarm, Poland) with methanol was carried out in a
3
1
00 cm glass reactor at atmospheric pressure following the proce-
1
.28 wt.%, linolenic 0.56 wt.%, and other acids 0.39 wt.%) were deter-
dure reported in our previous papers [32,34]. Reactor was equipped
with a reflux condenser, magnetic stirrer, and a tube for sam-
pling the solution. In catalytic experiment, triacetin or castor oil,
methanol and internal standard (toluene or eicosane) were intro-
duced to the reactor, heated up to a given temperature and then
mined in castor oil. and the average molecular weight of castor
oil was calculated to be 928 g/mol [34]. In discussion of catalytic
results, the yield of methyl esters formed in methanolysis of castor
oil was taken into consideration, identically to method widely used
in the case of vegetable oils. The yield of methyl esters formed in
methanolysis of castor oil was expressed in terms of the percentage
of methyl esters produced.
the catalyst was added.
For transesterification of triacetin 2.6 cm3 of triacetin and an
excess of methanol 16.2 cm3 were used. This composition corre-
As a measure of catalyst activity initial rate of methyl esters
sponds to typical molar ratio of methanol to triacetin (MR) equal
to 29 which was commonly used in our previous studies. The mix-
(ME) formation (below 10%) related to 1 g of catalyst was assumed
◦
[r (ME), mol ME/min g].
ture was heated to the desired temperature (50 C), and then 0.3 g
The catalytic tests were performed 2 times (in selected cases 3
times) and the average values are reported. The precision of chro-
matographic analysis was 5%.
of catalyst was added, (if not otherwise mentioned) and stirred
continuously. This amount of catalyst corresponds to catalysts con-
centration in the reaction mixture equal to 15.8 g/dm .
3

A. Zi e˛ ba et al. / Applied Catalysis A: General 387 (2010) 13–25
17
Table 1
Physicochemical characterization of catalysts properties.
2
3
+
Catalyst
Surface area [m /g]
Porosity [cm /g]
Acid capacity [mmol H /g]
Acid sites strength
(␣-TACT)
(␣-CAS)
(␣-GL)
PANI-S
25.9
41.6
164
1
136
4.7
0.054
0.24
0.56
4.75
4.70
0.15
0.32
0.16
0.32
1.5 < H0 < 0.8
12.1 ± 0.7
16.7 ± 2.3
22.1 ± 1.4
3.3 ± 0.1
7.3 ± 0.9
9.1 ± 0.2
49.4 ± 1.6
36.3 ± 1.7
48.9 ± 2.3
42.3 ± 1.5
41.9 ± 2.5
43.1 ± 1.3
53.5 ± 2.1
61.6 ± 3.4
61 ± 3.5
Amberlyst-15
Nafion SAC-13
Cs-2
Cs-2.5
Ag-2
0.8 < H0 < −3.0
−12 < H0 < −12.7
−13.2 < H0 <−13.8
−13.2 < H0 <−13.8
−3.0 < H0 < −8.2
37.7 ± 2.1
39.8 ± 1.4
24 ± 1.7
3
. Results and discussion
Physicochemical properties of studied catalysts are collected
physicochemical properties, among them capacity of acid cen-
tres and their strength determined in the present work for both
commercial resins, Amberlyst-15 and Nafion-SAC-13 agree well
with the data reported by other authors [18,46]. Acid capacity of
0.13–0.16 mmol/g was reported for Nafion-SAC-13 composite and
4.7 mmol/g for Amberlyst-15 resin [18,46].
In order to compare hydrophobic/hydrophilic nature of cata-
lysts, studies were undertaken to evaluate their affinity to highly
hydrophobic triglyceride reactant relative to that toward rather
hydrophilic, glycerol. These studies included the transesterification
experiments as well as physicochemical measurements.
in Table 1. Among polymeric catalysts, Nafion-SAC-13 composite
exhibits the highest specific surface area (164 m /g) and the highest
porosity because the polymer is dispersed throughout the matrix of
silica. Surface area and porosity of other two polymers, Amberlyst-
2
1
5 and PANI-S are definitively lower and porosity is especially low
in the case of PANI-S sample. In all polymeric catalysts a bimodal
distribution of the pores is observed and mesopores with average
pore diameters in the range of 9–18 nm dominate (Fig. 1). The SEM
image of PANI-S powder is displayed in Fig. 2. An aggregated gran-
ular morphology typical of polyanilines is observed. In catalytic
experiments the sieve fraction of particles of size 40–80 ␮m of
Amberlyst-15 and Nafion-SAC-13 polymers was used.
Cesium and silver salts of HPW, although exhibit crystalline
morphology in powdered state, they formed colloidal disper-
sion under conditions of triglyceride transesterification. Colloidal
particles existing in reaction mixture were determined to be
Three series of catalytic experiments were performed. In all
series, reaction was carried out under the same conditions. The
catalytic tests of series I are methanolysis experiments performed
with typical procedure described in Experimental. In series II exper-
iments, known amounts of glycerol (corresponding to 50% and 100%
of glycerol content formed in the experiment) were added to the
reaction mixture prior to the catalytic experiment. After heating
glycerol-containing reaction mixture up to the desired temper-
◦
3
00–600 nm in size [34]. As shown in Table 1, the catalysts essen-
ature (50 C) the catalyst was introduced and the reaction was
tially differ with respect to acid capacities. The concentrations of
acid sites are high and comparable for Amberlyst-15 and PANI-S
started. From these experiments, the affinity of catalysts toward
interaction with glycerol can be compared. In order to compare
the affinity of catalysts to triglyceride, series III experiments were
performed, in which the sample of catalyst was contacted with
(
4.70 and 4.75 mmol/g, respectively) and both capacities are dis-
tinctly higher than that of Nafion-SAC-13 composite (0.15 mmol/g).
The concentration of acid sites in the HPW salts, were assumed
to be the stoichiometric ones, according to the reports of other
authors [30]. The catalysts differ also with respect to the strength
of acid centres determined by Hammett indicators method. Among
all studied catalysts, PANI-S is the catalyst with the weakest acid
sites, the centres are stronger in Amberlyst-15 resin and much
stronger are the sites in Nafion-SAC-13 composite. In all three HPW
salts the acid centres are of higher strength than in polymeric cat-
alysts. However, acid centres of Ag-2 salt exhibit weaker acidity
compared to that of Cs salts which are known to be superacids
◦
triacetin (30 min, at 50 C) before methanolysis reaction. These
experiments were carried out in view of results reported by Liu et al.
[47]. When hydrotalcite catalysts were pretreated by triglycerides
prior to transesterification test, their activity dropped. According to
the authors, this negative role of pre-adsorbed triglycerides result-
ing in drop of catalysts activity was associated with the strong
adsorption of triglycerides on active sites [47].
The affinity of catalyst to non-polar triglyceride relative to that
toward polar glycerol was also evaluated from the measurements of
contact angles for droplets of triacetin (␣-TACT), castor oil (␣-CAS)
and glycerol (␣-GL) on the surface of pressed tablet of catalysts
(Table 1). It should be noticed that in limiting case, if the liquid is
very strongly attracted to the solid surface, the droplet will com-
pletely spread out on the solid surface and the contact angle will
[
28,43]. The relation of acid sites strength determined for Ag and
Cs salts agrees with the report of Moffat and coworker [45]. The
◦
be close to 0 .
The measured values of castor oil contact angles (␣-CAS) are of
◦
similar order for all studied catalysts, 36–49 and all are relatively
low showing that droplets of castor oil are attracted to the sur-
face of all catalysts (Table 1). This shows hydrophobic character of
all studied catalysts. All studied catalysts exhibit also low triacetin
◦
contact angles (␣-TACT) within the range 3–22 showing that the
droplets of triacetin are well attracted to their surface. However, the
values of triacetin contact angles (␣-TACT) differ showing that the
affinity of catalysts toward triacetin molecules varies. The strongest
◦
attraction of triacetin is observed for Cs salts (␣-TACT = 3.3 ) the
catalysts with the strongest acid sites. It should be stressed that
hydrophobic character of Cs-2.5 salt observed in the present work is
consistent with the data of Okuhara and coworker [48] who termed
this salt as “hydrophobic” solid catalyst from the measurements of
its interactions with benzene. The attraction of triacetin is weaker
in case of PANI-S and Amberlyst-15 polymers with weaker acid
◦
sites. The value of (␣-TACT) for Nafion-SAC-13 is the highest, 22 ,
Fig. 2. SEM image of PANI-S.

1
8
A. Zi e˛ ba et al. / Applied Catalysis A: General 387 (2010) 13–25
Fig. 3. Methanolysis of triacetin catalysed by soluble H2SO4 and H3PW12O40 acids. Reagents distribution profiles and selectivity to glycerol (GL), diacetin (DI) and monoacetin
◦
(
MONO) (reaction conditions: MR = 29, temp. 50 C).
showing that triacetin is slightly weaker attracted to the surface
of this composite maybe owing to the more polar Si–OH groups
present in this composite. It seems from the obtained data that
in the spreading of the droplets of triacetin, apart from “typical
hydrophobic/hydrophilic interactions” also the acidity of catalyst
surface playedarole. On the otherhand, the castor oil contact angles
PANI-S was associated with the presence of sulfuric acid [25,26]
whereas in Cs and Ag salts of HPW, protons are the active cen-
ters [28,33]. Therefore, methanolysis experiments were performed
in the presence of soluble H SO and H PW12O40 acids (dried at
2
4
3
◦
230 C before experiment [32,49]). Product distribution profiles
obtained are displayed in Fig. 3. They differ with respect to the
course of consecutive reactions transforming triacetin via diacetin
and monoacetin to glycerol. Similar difference was also observed
upon transesterification of rapeseed oil catalysed by various soluble
Keggin heteropolyacids and sulfuric acid, however, no comments
were provided by the authors [49]. As Fig. 3 shows in the presence
of sulfuric acid both partial glycerides, diacetin and monoacetin
were formed from the very beginning of reaction and their contents
passed through the maximum. The maximum content of diacetin
is low 0.087 mol/dm3 (corresponds to diacetin yield of 12%) and
is reached at ca. 40% triacetin conversion whereas the maximum
yield of monoacetin is very high. A different effect can be seen in
the presence of soluble HPW. From the very beginning of reac-
tion the only partial glyceride formed is diacetin which attained
high maximum content, 0.112 mol/dm3 (corresponds to diacetin
yield of 15.5%) whereas, monoacetin is observed only at higher tri-
acetin conversion, above 50% and its maximum content is very low
(Fig. 3). This difference can be easier recognized from the graphs
(
␣-CAS) are higher compared to the values of (␣-TACT) and they do
not allow determination of difference between properties of indi-
vidual catalysts. This may be associated with high viscosity of castor
oil (ca. 10 times higher compared to triacetin) and the presence of
triglycerides consisting of C–OH groups.
As shown in Table 1, the catalysts strongly differ with respect
to the glycerol contact angles (␣-GL). For polymeric catalysts,
the values of (␣-GL) all are comparable and within the range
◦
3.5–61.6 whereas all three salts of HPW, Ag-2, Cs-2 and Cs-2.5
5
◦
◦
◦
have definitively lower glycerol contact angles, 24 , 37.7 and 39.8 ,
respectively. This shows that, glycerol is well attracted to the sur-
face of all HPW salts, definitively better than to the surface of
polymeric catalysts. Thus, all studied catalysts exhibit much better
affinity to non-polar triacetin molecules than to the polar reagent,
glycerol.
The activity of Amberlyst-15 and Nafion-SAC-13 catalysts was
reported to be due to the acid centers, SO H [5] and that of
3

A. Zi e˛ ba et al. / Applied Catalysis A: General 387 (2010) 13–25
19
Fig. 4. Transesterification of triacetin in the presence of PANI-S, Amberlyst-15,
Fig. 5. Transesterification of triacetin in the presence of Cs-2 and Ag-2 salts at cat-
◦
Nafion-SAC-13, Cs-2 and Ag-2 catalysts (reaction conditions: MR = 29, temp. 50 C,
3
3
alyst concentration of 7 g/dm (open marks) and 15.8 g/dm (full marks) (reaction
3
catalysts concentration 15.8 g/dm ).
◦
conditions: MR = 29, temp. 50 C).
reporting the selectivity to glycerol and partial glycerides. From
the very beginning of reaction up to 30–40% triacetin conversion,
the selectivity to glycerol is as high as 65% over HPW whereas the
selectivity is lower ca. 50% in sulfuric acid catalysed transester-
ification. An opposite relation is observed for selectivity toward
diacetin. Thus, from the very beginning of transesterification catal-
ysed by soluble HPW, the fraction of triacetin reacted to glycerol
was higher than in the presence of sulfuric acid. This effect may
be related to some of properties of studied acids and in particu-
lar higher strength of acid protons in H PW12O40 may be taken
3
into account [28]. It has also been reported that due to higher acid
strength of HPW at equivalent H+ concentrations HPW-catalysed
transesterificationachieveshigherreactionrates than conventional
mineral acids (H SO , H PO ) [32,49].
2
4
3
4
Owing to a difference in performance of H SO4 and H PW12O40
2
3
acids, the results obtained on polymeric catalysts are compared
with that for H SO and catalytic performance of Ag and Cs salts is
2
4
compared with that of soluble H PW12O40
.
3
3.1. Series I experiments
Fig. 4 shows the time conversion plots for transesterification
of triacetin. All catalysts are effective for the methanolysis of tri-
acetin. As a measure of catalyst activity, initial rate of triacetin
conversion (below 10% conversion) per unit mass of catalyst was
assumed (Table 2). Taking into consideration initial rate per unit
Fig. 6. The change of triacetin concentration in methanolysis catalysed by Cs-2, Cs-
◦
2
7
.5 and Ag-2 salts (reaction conditions: MR = 29, temp. 50 C, catalysts concentration
g/dm ).
3
Table 2
Initial activity of catalysts in transesterification of triacetin and castor oil and the drop of initial rate induced by glycerol and triacetin.
−
1
−1
g ]
Catalyst
Maximum
conversion
Maximum
yield [%]
R
Activity drop [%]
r [mol(ME) min
10
−
1
−1
−4
[mol(TACT) min
g
]
−
10
4
Triacetin pre-treatment
Glycerol (50%)
Glycerol (100%)
[%]
PANI-S
99
96.4
96.5
70.4a
91.1
1.76
2.26
0.61
7.28
4.0
21
19
35
45
46
8
9
39
69
75
13
22
70
83
87
58
0.120
0.125
0.035
0.420
0.088
0.760
Amberlyst-15
Nafion-SAC-13
Cs-2
Cs-2.5
Ag-2
98.8
86.2a
94.6
92.6
94.1
5.10
a
Determined after 30 h of reaction.

2
0
A. Zi e˛ ba et al. / Applied Catalysis A: General 387 (2010) 13–25
Fig. 7. Transesterification of triacetin catalysed by PANI-S, Amberlyst-15, Nafion-SAC-13 and soluble H2SO4. The yield of diacetin (a) and monoactin (b) vs. conversion of
◦
3
triacetin, selectivity to glycerol (GL), diacetin (DI) and monoacetin (MONO); (c and d) (reaction conditions: MR = 29, temp. 50 C, solid catalyst concentration 15.8 g/dm ).
mass of catalysts, Cs and Ag salts of HPW exhibited superior initial
activity compared to all polymeric catalysts and the most active are
Cs salts. Among polymeric catalysts the most active is Amberlyst-
catalysts. Initially high activity of Cs-2 salt slowly decreases when
ca. 20–30% of TACT was reacted and reaction progresses further at
distinctly lower rate. This difference can be better recognized from
graphs displayed in Fig. 5. The time conversion plots obtained at
two different concentrations of Cs-2 salt in reaction mixture are
compared with the results obtained at the same concentrations
of Ag-2 salt. A difference in performance of Ag-2 and Cs-2 salts
can be seen at both concentrations of catalysts. Analogous effect,
decrease of initially high activity is also observed for Cs-2.5 salt. The
decrease of triacetin concentration vs. time of reaction carried out
in the presence of Cs-2, Cs-2.5 and Ag-2 salts is compared in Fig. 6.
An inhibition of triacetin conversion is observed on both Cs-2 and
Cs-2.5 salts, which makes their performance different from that of
Ag-2 salt.
It should be noticed that although initially high rate on Cs salts
decreases and further reaction progresses at lower rate, high con-
version of triacetin (94.6%) was obtained on these salts. However,
at the same reaction conditions this conversion on Cs-2 salt was
reached after 15 h of reaction, i.e. 5 times longer reaction time com-
pared to Ag-2 salt (3 h). This may suggest a partial blockage of active
sites in the Cs salts during the transesterification of triacetin. This
deactivation behaviour is observed for Cs salts only.
1
5 resin. The initial rate of triacetin conversion in the presence of
−1
−1
g
Amberlyst-15 (2.26 mol min
rate on PANI-S (1.76 mol min
) is only slightly higher than the
−1
g ) but both rates are distinctly
−1
higher than the rate in the presence of Nafion-SAC-13 composite.
It should be remembered that catalyst activity is a func-
tion of both acid site strength and concentration. Amberlyst-15
and PANI-S catalysts, exhibit high and comparable concen-
trations of acid centres, but these sites are weak. On the
other hand, the concentration of acid centres in Nafion-SAC-
1
3 is distinctly lower (0.15 mmol/g) but these centres are of
higher strength. This may explain the observed order of activ-
ity.
It can be observed that although the concentrations of acid sites
in Nafion-SAC-13 and Cs-2.5 are similar (0.15 and 0.16 mmol/g,
respectively) performance of Cs-2.5 differs significantly from that of
Nafion-SAC-13. This may be related to the strength of sites, which
are stronger in Cs-2.5 salt. Furthermore, it can be observed from
the time conversion plots (Fig. 4) that catalytic performance of Cs-
2
salt remarkably differs from that of Ag-2 salt and all other studied

A. Zi e˛ ba et al. / Applied Catalysis A: General 387 (2010) 13–25
21
Fig. 8. Transesterification of triacetin catalysed by Cs-2, Cs-2.5, Ag-2 and soluble H3PW12O40. The yield of diacetin and monoactin (a) vs. conversion of triacetin, selectivity
◦
3
to glycerol (GL), diacetin (DI) and monoacetin (MONO) (b and c) (reaction conditions: MR = 29, temp. 50 C, concentration of HPW salts 15.8 g/dm ).
In Fig. 7 the yields of partial glycerides (diacetin, monoacetin)
are plotted vs. the conversion of triacetin for reactions performed
in the presence of polymeric catalysts and soluble H SO . In the
of all three polymeric catalysts. This may be explained by an influ-
ence of hydrophobic environment of active centers in polymeric
catalysts which in agreement with Liu et al. [36] can make more
difficult further transformation of more polar molecules of partial
glycerides, diacetin and monoacetin to glycerol.
In Fig. 8 the yields of partial glycerides obtained using Ag
and Cs salts are compared with those in the presence of soluble
2
4
presence of all three polymeric catalysts the yield of diacetin is
higher and that of monoacetin is lower compared to those under
homogenous conditions using sulfuric acid. The selectivity data
towards glycerol and partial glycerides are compared in Fig. 7c.
Selectivity changes over Amberlyst-15 and PANI-S catalysts are
of similar character to that for soluble sulfuric acid. In all three
reaction systems at triacetin conversion below 30–40%, selectivity
to glycerol achieves 50–60% and slowly increases as the reaction
progresses. The course of consecutive reactions on Nafion-SAC-13
composite differs to some extent because selectivity to glycerol
is lower (ca. 35%) and that to diacetin is higher. This shows that
the course of consecutive reactions leading to glycerol is more
restricted on Nafion-SAC-13 than on other two polymers maybe
because of higher strength of acid sites and/or the presence of
silanol groups in the composites. These results show that the trans-
formation of triacetin via intermediates, diacetin and monoacetin
to glycerol on solid catalysts changes relative to soluble H SO . In
H PW12O40. In all reaction systems diacetin is formed from the very
3
beginning of reaction. The maximum yield of diacetin on Ag-2 salt
is higher compared to soluble HPW whereas on Cs-2 salt the maxi-
mum yields are comparable. However, the growth of diacetin yield
relative to soluble HPW is smaller than in case of polymeric PANI-S
and Nafion-SAC-13 catalysts. In the presence of Cs salts the con-
tent of diacetin grows along to “S-shape curve”, which differs their
performance from that of HPW and Ag-2 salt. A difference can also
be observed in the yield of monoacetin. In the presence of Cs salts,
monoacetin is formed at low triacetin conversion ca. 30%, which is
definitively lower compared to soluble HPW (ca. 50%). In contrast,
on Ag-2 salt monoacetin is formed at very high triacetin conver-
sion. This difference in performance of Ag-2 and Cs salts resulted
in quite different changes in selectivity to glycerol.
2
4
particular the higher yields of diacetin are observed in the presence

2
2
A. Zi e˛ ba et al. / Applied Catalysis A: General 387 (2010) 13–25
Fig. 9. The effect of glycerol and triacetin-pretreatment on the yield of diacetin and monoacetin, reactions catalysed by PANI-S, Amberlyst-15, Nafion-SAC-13 (reaction
◦
3
conditions: MR = 29, temp. 50 C, catalyst concentration 15.8 g/dm ).
As shown in Fig. 8b, from the very beginning of reaction on Ag-2
salt selectivity to glycerol is ca. 50% and it changes in similar man-
ner to that on PANI-S and Amberlyst-15 polymers. On the other
hand on Cs salts, selectivity to glycerol is very high, ca. 80%, higher
than in the presence of soluble HPW. However, when ca. 20–25%
of triacetin was reacted, selectivity to glycerol slowly decreases
maybe due to the growing content of monoacetin. These effects
may suggest strong interaction of partial glycerides/glycerol with
active centers of Cs salts. This interaction may facilitate the trans-
formation of diacetin to monoacetin as well it may facilitate the
reverse reaction, transformation of diacetin to monoacetin (Fig. 8a).
On the other hand, very high initial activity of Cs-2 salt remark-
ably decreased when ca. 20–25% of triacetin was reacted (Fig. 4).
Thus, strong interaction between active centers (protons) of Cs salts
and molecules of glycerol/partial glycerides may be considered as
the reason of partial deactivation of Cs salts which started already
at low conversion of triacetin, ca. 20–25%. This type behavior was
not observed for Ag-2 salt and all other studied catalysts. Com-
parison of reagents distribution profiles on Ag-2 and polymeric
catalysts (Fig. 7) with those on Cs salts (Fig. 8) show a difference
in the course of consecutive reactions. In the former case, the yield
of diacetin is higher and selectivity to glycerol is lower compared
to Cs salts. This shows that on Cs salts the conversion of triacetin
via di- and mono-glycerides to final product, glycerol is easier and
the course of consecutive reactions is less inhibited. As already
described, this inhibition leading to growing content of diacetin
was to some extent influenced by hydrophobic properties of cata-
lyst. Hence, the presence of acid sites of high strength, as in the Cs
salts, seems to be more decisive than hydrophobicity of the nearest
environment of these acid sites.
Now, the question arises as to whether the interaction of cata-
lyst with glycerol impacts their activity. To verify this effect, series
II experiments were performed in which reaction was carried out
in the presence of glycerol. As described before, known amounts
of glycerol (corresponding to 50% and 100% of glycerol content
which can be formed in the experiment) were added to the reac-
tion mixture before the catalytic experiment. In Table 2, the drop
of initial rate (in %) relative to the rate obtained under typical pro-
cedure of catalytic test (series I) is reported. It is observed that
added glycerol exerts only slightly the activity of polymeric cat-
alysts, Amberlyst-15 and PANI-S because their activity dropped
by ca. 8–9% only. In the presence of added glycerol the course of

A. Zi e˛ ba et al. / Applied Catalysis A: General 387 (2010) 13–25
23
Fig. 10. The yield of diacetin in methanolysis catalysed by Cs-2, Cs-2.5 and Ag-2 salts. The role of glycerol and triacetin pre-treatment (reaction conditions: MR = 29, temp.
◦
3
5
0 C, catalyst concentration 15.8 g/dm ).
consecutive reactions on PANI-S does not essentially change, evi-
denced by practically the same yield of diacetin. The changes are
slightly higher on Amberlyst-15 resin (Fig. 9). The drop of activity
was higher for Nafion-SAC-13 catalyst (by 39%) and the growth of
diacetin yield was higher. Deactivation was the strongest for Cs-2
and Cs-2.5 salts, their activity dropped by 69% and 75%, respec-
tively. In their presence, conversion of triacetin was suppressed
from the very beginning of reaction [50]. Furthermore, as shown
in Fig. 10, the yield of diacetin was much higher than in reaction
without added glycerol.
active centers of these catalysts exhibited the highest strength. This
resulted in the strongest deactivation of Cs salts. On the other hand,
although glycerol was well attracted to the surface of Ag-2 salt, its
acid centers are weaker and deactivation due to added glycerol was
smaller. Thus, acid centers of high strength may be considered as
factor facilitating interaction with polar reagents such as glycerol.
Similar effect, strong interaction with active centers and accu-
mulation of partial glycerides/glycerol was observed on alkaline
Mg-Al hydrotalcites [51] as well as on alkali supported on CaO
and MgO catalysts [52]. The authors concluded that accumulation
of polar reagents (glycerides/glycerol) could result in the creation
of glycerol film onto the solid catalyst. The presence of this glyc-
erol/glycerides film could have two important effects. It could
present a mass transfer barrier to the methanol and triglycerides.
Furthermore, it could facilitate the reverse reactions thus shifting
reactions to left, to the intermediate products i.e. partial glycerides
[52]. This type of effects seems to be very probable also in the
presence of Cs salts.
The observed deactivation of catalysts due to added glyc-
erol correlates with the strength of acid sites and the affinity of
catalysts toward glycerol which was characterized by glycerol con-
tact angles. The effect of glycerol is the smallest for PANI-S and
Amberlyst-15 catalysts. Weak attraction of glycerol to the surface
of such two polymers (␣-GL equal to 61.6 and 61, respectively)
is observed and both catalysts have weak acid centers. Although
◦
Nafion-SAC-13 have similar ␣-GL = 61 , its acid centers are stronger
and deactivation due to added glycerol was higher.
The effect of added glycerol was the strongest for Cs salts. Glyc-
erol was very well attracted to the surface of Cs salts and the
Series III experiments were performed in the presence of cata-
lysts pretreated with triacetin, i.e. non-polar reagent. The triacetin
pretreatment was carried out at temperature of reaction, 50 C.
◦

2
4
A. Zi e˛ ba et al. / Applied Catalysis A: General 387 (2010) 13–25
In conclusions, all studied catalysts exhibit better affinity
toward highly hydrophobic triglycerides reactants compared to
that toward rather hydrophilic glycerol. The affinity of catalysts
to glycerol differs being the highest for the Cs and Ag salts of
HPW. Therefore affinity of studied catalysts toward glycerol and the
strength of acid sites seem to be decisive factors for their activity
and in particular for deactivation due to polar reagents adsorption.
Greater affinity toward glycerol and higher strength of acid sites
will result in easier deactivation of catalysts. However, the com-
parison of results for Ag-2 and Cs-2 salts shows that the strength of
acid sites seems to be more important than the affinity toward glyc-
erol. This conclusion is supported by the observation that glycerol
was better attracted to the surface of Ag-2 salts compared to Cs-
2
, however, due to weaker acid sites deactivation of Ag-2 salt was
negligible. The results obtained in this work well support literature
suggestions/observations that a moderate strength of acid sites and
their hydrophobic microenvironment should promote the activity
and stability of solid acid catalysts in reaction mixture consisting
of methanol and triglyceride [5,24].
4. Conclusions
Transesterification of triacetin with methanol was studied
Fig. 11. The yield of methyl esters against reaction time, methanolysis of castor oil,
reaction conditions: MR = 29, temp. 60 C, catalyst concentration 43 g/dm ).
◦
3
in the presence of solid acid catalysts Amberlyst-15, Nafion-
SAC-13, polyaniline-sulfate, Ag HPW12O40, Cs HPW12O40 and
2
2
Cs2.5H0.5PW12O40. The initial activity per unit mass of catalyst in
The activity of all triacetin-pretreated catalysts dropped but to
various extents (Table 2). Similar effect, the drop of activity was
reported by Liu et al. [47]. It is observed that deactivation was low
for PANI-S and Amberlyst-15 catalysts, 21% and 19%, respectively.
Deactivation of Nafion-SAC-13 composite was stronger (ca. 35%)
and deactivation was the strongest in case of Cs salts. Their activity
dropped by ca. 45%. An influence of triacetin pretreatment on the
yield of diacetin is reported in Figs. 9 and 10. The yield of diacetin
practically does not change in the presence of triacetin-pretreated
PANI-S and Amberlyst-15 catalysts, grows only slightly in the pres-
ence of Nafion-SAC-13 composite. The increase of diacetin yield
is the highest on Cs-2 and Cs-2.5 salts however, this increase is
lower compared to that induced by added glycerol (Fig. 10). Hence,
the obtained order of activity drop correlates with the strength of
acid sites determined by Hammett indicators for studied catalysts.
An exception is Nafion-SAC-13 composites, maybe because of the
presence of more polar silanol groups.
reaction of triacetin was the highest for Cs salts, Cs HPW12O40 and
2
Cs2.5H0.5PW12O40 however, the catalysts partially deactivated in
the course of reaction. The highest strength of acid sites and high
affinity toward polar glycerol was determined for these Cs salt.
The affinity of catalyst to highly hydrophobic triglyceride reactant
relative to that toward rather hydrophilic glycerol was found to
affect the course of consecutive reactions transforming triacetin via
partial glycerides (diacetin, monoacetin) to final product, glycerol.
Furthermore, affinity of studied catalysts toward glycerol and the
strength of acid sites seem to be decisive factors for their activity
and in particular for deactivation due to polar reagents adsorption.
Greater affinity toward glycerol and higher strength of acid sites,
deactivation of catalysts would be easier. The results obtained in
this work well support literature suggestions/observations that a
moderate strength of acid sites and their hydrophobic microen-
vironment should promote the activity and stability of solid acid
catalysts in transesterification of triglycerides.
Let us compare the activity of studied catalysts in the transester-
ification of vegetable oil, castor oil. As a measure of catalyst activity
Acknowledgements
−
1
−1
initial rate of methyl ester formation (r (ME), mol min
g ) was
assumed (Table 2). The yield of methyl esters against reaction time
is shown in Fig. 11. For polymeric catalysts the order of activity is the
same as that observed in methanolysis of triacetin. The activities of
Amberlyst-15 and PANI-S are higher than that of Nafion-SAC-13.
Moreover, activity of Nafion-SAC-13 is the lowest, ca. 3.7 times
lower than that of Amberlyst-15 (3.6 times lower in methanol-
ysis of triacetin). In contrast to polymeric catalysts, the activity
relations changed for Cs, Ag salts of HPW (Fig. 11). In the pres-
ence of Ag-2 salt methanolysis of vegetable oil is faster reaction
than on Cs-2 and Cs-2.5 salts. In addition, the activity of Cs-2 salt
is ca. 4 times higher than that of Cs-2.5 whereas in methanolysis
of triacetin Cs-2 was only 1.8 times more active than Cs-2.5 salt.
This may be explained taking into account that the accessibility
of large triglyceride molecules present in castor oil to the active
sites is more difficult and the activity under these conditions may
be mainly related to the centers in outermost surface of catalysts
particles. Under these conditions the centers may be deactivated
due to glycerol already from the very beginning of the transesteri-
fication. As a consequence, activity of Cs salts is low from the very
beginning of castor oil transesterification.
A. Zi e˛ ba acknowledges a Ph.D. grant of the Polish Academy of
Sciences. Thanks are also to the Grant Agency of the Academy of
Sciences of the Czech Republic (IAA 400500905).
References
[1] E. Lotero, Y. Liu, D.E. Lopez, K. Suwannakarn, D.A. Bruce, J.G. Goodwin Jr., Ind.
Eng. Chem. Res. 44 (2005) 5353–5363.
[
2] M. Zabeti, W.M.A. Wan Daud, M. Kheireddine Aroua, Fuel Process. Technol. 90
2009) 770–777.
(
[3] M. Di Serio, R. Tesser, L. Pengmei, E. Santacesaria, Energy Fuels 22 (2008)
207–218.
[
4] Z. Helwani, M.R. Othman, A. Aziz, J. Kim, W.J.N. Fernando, Appl. Catal. A 363
2009) 1–10.
(
[5] J.A. Melero, J. Iglesias, G. Morales, Green Chem. 11 (2009) 1285–1308.
[6] J. Jitputti, B. Kitiyanan, P. Rangsunvigit, K. Bunyakiat, L. Attanatho, P. Jenvanit-
panjakul, Chem. Eng. J. 116 (2006) 61–66.
[
7] C. Martins Garcia, S. Teixeira, L. Ledo Marciniuk, U. Schuchardt, Bioresour. Tech-
nol. 99 (2008) 6608–6613.
[8] D.E. Lopez, J.G. Goodwin Jr., D.A. Bruce, S. Furuta, Appl. Catal. A 339 (2008)
6–83.
9] K. Suwannakarn, E. Lotero, J.G. Goodwin Jr., C. Lu, J. Catal. 255 (2008) 279–286.
10] S. Furuta, H. Matsuhashi, K. Arata, Catal. Commun. 5 (2004) 721–723.
[11] K. Jackobson, R. Gopinath, L.C. Meher, A.K. Dalai, Appl. Catal. B 85 (2008) 86–91.
7
[
[

A. Zi e˛ ba et al. / Applied Catalysis A: General 387 (2010) 13–25
25
[
12] M.G. Kulkarni, R. Gopinath, L.C. Meher, A.K. Dalai, Green Chem. 8 (2006)
[34] A. Zieba, L. Matachowski, J. Gurgul, E. Biela n´ ska, A. Drelinkiewicz, J. Mol. Catal.
A 316 (2010) 30–44.
1
056–1062.
[
[
13] L. Xu, Y. Wang, X. Yang, J. Hu, W. Li, Y. Guo, Green Chem. 11 (2009) 314–317.
14] F. Chai, F. Cao, F. Zhai, Y. Chen, X. Wang, Z. Su, Adv. Synth. Catal. 349 (2007)
[35] D.E. Lopez, J.G. Goodwin Jr., D.A. Bruce, E. Lotero, Appl. Catal. A: Gen. 295 (2005)
97–105.
1
057–1062.
15] M.H.Z. Zong, Q. Duan, W.Y. Lou, T.J. Smith, H. Wu, Green Chem. 9 (2007)
34–437.
[36] Y. Liu, E. Lotero, J.G. Goodwin Jr., Ch. Lu, J. Catal. 246 (2007) 428–433.
[37] Y. Du, S. Liu, Y. Ji, Y. Zhang, S. Wei, F. Liu, F.-S. Xiao, Catal. Lett. 124 (2008) 133.
[38] G.S. Macala, A.W.Ch. Robertson, L. Johnson, Z.B. Day, R.S. Lewis, M.G. White, A.V.
Iretskii, P.C. Ford, Catal. Lett. 122 (2008) 205–209.
[
4
[
[
[
[
[
16] X. Mo, E. Lotero, Ch. Lu, Y. Liu, J.G. Goodwin Jr., Catal Lett. 123 (2008) 1–6.
17] M.A. Harmer, Q. Sun, Catal. Appl. A 221 (2001) 45–62.
[39] L. Zhang, W. Guo, D. Liu, J. Yao, L. Ji, N. Xu, E. Min, Energy Fuels 22 (2008)
1353–1357.
18] D.E. Lopez, J.G. Goodwin Jr., D.A. Bruce, Catal. J. 245 (2007) 381–391.
19] B.M.E. Russbueldt, W.F. Hoelderich, Catal. Appl. A 362 (2009) 47–57.
20] S.M. de Rezende, M. de Castro Reis, M.G. Reid, P.L. da Silva Jr., F.M.B. Coutinho,
R.A. da Silva San Gil, E.R. Lachter, Appl. Catal. A, 349 (2008) 198–203.
21] I.K. Mbaraka, B.H. Shanks, JAOCS 83 (2006) 79–91.
[40] S.M. Plentz Meneghetti, M.R. Menghetti, C.R. Wolf, E.C. Silva, G.E.S. Lima, M.A.
de Coimbra, J.I. Soletti, S.H.V. Carvalho, JAOCS 83 (2006) 819–825.
[41] S.M. Plentz Meneghetti, M.R. Menghetti, C.R. Wolf, E.C. Silva, G.E.S. Lima,
L. deLira Silva, T.M. Serra, F. Cauduro, de Oliviera, Energy Fuels 20 (2006)
2262–2269.
[42] J. Haber, L. Matachowski, D. Mucha, J. Stoch, P. Sarv, Inorg. Chem. 44 (2005)
6695–6703.
[43] T. Okuhara, T. Nishimura, H. Watanabe, M. Missono, J. Mol. Catal. 74 (1992)
247–256.
[
[
[
22] L. Xu, W. Li, J. Hu, X. Yang, Y. Guo, Appl. Catal. B 90 (2009) 587–594.
23] I.K. Mbaraka, K.J. McGuire, B.H. Shanks, Ind. Eng. Chem. Res. 45 (2006)
3
022–3028.
24] J.A. Melero, L.F. Bautista, G. Morales, J. Iglesias, D. Briones, Energy Fuels 23
2009) 539–547.
[
(
[
[
25] S. Palaniappan, M. Sai Ram, Green Chem. 4 (2002) 53–55.
26] S. Palaniappan, C. Saravanan, C.A. Amarnath, V.J. Rao, Catal. Lett. 97 (2004)
[44] A.K. Singh, S.D. Fernando, Energy Fuels 22 (2008) 2067–2069.
[45] M.A. Parent, J.B. Moffat, Catal. Lett. 48 (1997) 135–143.
[46] M.A. Harmer, Q. Sun, Appl. Catal. A 221 (2001) 45–55.
[47] Y. Liu, E. Lotero, J.G. Goodwin Jr., X. Mo, Appl. Catal. A 331 (2007) 138–148.
[48] T. Yamada, T. Okuhara, Langmuir 16 (2000) 2321–2325.
[49] P. Morin, B. Hamad, G. Sapaly, M.G. Carneiro Rocha, P.G. Pries de Oliviera, W.A.
Gonzalez, E. Andrade Sales, N. Essayem, Appl. Catal. A 330 (2007) 69–76.
[50] A. Matachowski, A. Zi e˛ ba, M. Zembala, A. Drelinkiewicz, Catal. Lett. 133 (2009)
49–62.
7
7–81.
[
27] A.J.T. Zieba, A. Drelinkiewicz, E.N. Konyushenko, J. Stejskal, Appl. Catal. A,
doi:10.1016/j.apcata.2010.05.042.
[
[
[
[
28] T. Okuhara, N. Mizuno, M. Misono, Adv. Catal. 41 (1996) 113–252.
29] A. Corma, A. Martinez, C. Martinez, J. Catal. 164 (1996) 422–432.
30] T. Okuhara, M. Kimura, T. Kawai, Z. Xu, T. Nakato, Catal. Today 45 (1998) 73–77.
31] K. Narasimharao, D.R. Brown, A.F. Lee, A.D. Newman, P.F. Siril, S.J. Tavener, K.
Wilson, J. Catal. 248 (2007) 226–234.
[51] D.G. Cantrell, L.J. Gillie, A.F. Lee, K. Wilson, Appl. Catal.
183–190.
A 287 (2005)
[
[
32] A. Zi e˛ ba, L. Matachowski, E. Lalik, A. Drelinkiewicz, Catal. Lett. 127 (2009)
1
83–194.
33] Y. Izumi, M. Ono, M. Kitagawa, M. Yoshida, K. Urabe, Microp. Mater. 5 (1995)
55–262.
[52] C.S. MacLeod, A.P. Harvey, A.F. Lee, K. Wilson, Chem. Eng. J. 135 (2008)
63–70.
2