Arabian Journal for Science and Engineering (2018) 43:5771–5783

https://doi.org/10.1007/s13369-017-3009-x

RESEARCH ARTICLE - CHEMICAL ENGINEERING

Caesium Salt of Tungstophosphoric Acid Supported on Mesoporous

SBA-15 Catalyst for Selective Esterification of Lauric Acid with Glycerol
to Monolaurin
Syamima Nasrin Mohamed Saleh1 · Mohd Hizami Mohd Yusoff1 · Ahmad Zuhairi Abdullah1

Received: 12 July 2017 / Accepted: 3 December 2017 / Published online: 11 December 2017
© King Fahd University of Petroleum & Minerals 2017

Abstract
Cs2.5 H0.5 PW12 O40 supported on SBA-15 catalysts were prepared via two-sequential-step post-impregnation method with
different ratios of cesium salt. The synthesized catalysts were then characterized using nitrogen adsorption–desorption, FTIR,
EDX, SEM, BET and TGA analyses in order to determine the physicochemical properties of the catalysts. The activity of the
catalysts in the esterification of glycerol to monolaurin was investigated under various reaction parameters including catalyst
loadings, reaction temperatures and glycerol-to-lauric acid molar ratio. The highest lauric acid conversion (71.8%) with 44.9%
of monolaurin yield was obtained using 20 wt%Cs-HPW/SBA-15 catalyst in 4 h at 170 ◦ C using 4:1 of glycerol-to-lauric acid
molar ratio and 2.5 wt% of catalyst loading. This catalyst was stable and reusable for up to three cycles in the esterification
reaction without significant loss in catalytic activity.

Keywords 12-Tungstophosphoric acid · SBA-15 · Selective esterification · Lauric acid · Monolaurin

1 Introduction effect with 70% of lauric acid conversion and 50% of mono-
laurin yield. However, the thermal stability and reusability
Recently, functionalized mesoporous silicas have attracted of the catalyst were not explored.
considerable interest in the production of monoglycerides Despite its good catalytic activity, leaching of the active
due to their uniform and large pore sizes which are advanta- sites from the SBA-15 during the reaction has become the
geous in facilitating the reaction involving bulky reactants. major concern for this catalyst. Thus, it becomes a challenge
Several attempts have been made on the functionalization to create a practical, durable and highly reactive catalyst,
of SBA-15 with various active species to increase the cat- which needs further modifications for the acidic site of the
alytic activity toward high monoglyceride selectivity. Other catalyst. This can be done by introducing the caesium (Cs)
than sulfated zirconia [1] and propyl sulfonic acid [2,3] as salts of tungstophosphoric acid (Cs-HPW) onto SBA-15.
active species, heteropolyacid (HPA) also has the potential Cs-HPW is known for being highly acidic and has unique
for such reaction. In a recent study, Simsek et al. [4] found physical characteristics and good activity for various organic
that STA/SBA-15 catalyst exhibited high monolaurin selec- reactions [6]. Therefore, by impregnating the Cs-HPW salt
tivity within 3 h of reaction but as the reaction prolonged, on the mesoporous SBA-15 support, the thermal stability
the selectivity was found to decrease. Meanwhile, Hoo et and the surface acidity of the catalyst should be improved
al. [5] synthesized 12-tungstophosphoric acid (HPW) sup- as well. Previously, Dias et al. [6] proved that the acidity
ported on SBA-15 catalyst for selective esterification of lauric was increased sharply as the Cs content increased and all Cs
acid with glycerol to monolaurin. After 6 h, the catalyst with derivatives of HPW revealed both Brønsted and hydrogen-
20 wt% HPW loading exhibited the highest shape selective bonded sites in the structure. Madhusudhan et al. [7] prepared
Cs salts of HPW deposited on SBA-15 and the catalyst
B Ahmad Zuhairi Abdullah showed considerably good performance in acidic reactions
[email protected] (propionylation and alkylation). In another study, Landau et
1 al. [8] also found that Cs salt of HPW dispersed on SBA-15
School of Chemical Engineering, Engineering Campus,
Universiti Sains Malaysia, 14300 Nibong Tebal, Penang, showed rather improved performance toward NOx storage.
Malaysia Although Cs salts of HPAs have shown good performances

123
5772 Arabian Journal for Science and Engineering (2018) 43:5771–5783

in various types of reactions, no work has been reported so impregnation method [11]. In a typical synthesis, 0.4 g of
far on the use of Cs salts of HPW supported on SBA-15 in 12-tungstophosphoric acid (HPW) was dissolved in 10 ml
selective esterification of lauric acid with glycerol to mono- of ethanol and the mixture was added into 1.6 g of SBA-15.
laurin. Then, the resultant product was dried at 60 ◦ C for 12 h and
In the present work, Cs-HPW acidic salt supported on then followed by another 12 h at 100 ◦ C.
SBA-15 catalysts has been synthesized and used in the ester-
ification of glycerol with lauric acid to monolaurin. Particular
attention has been given to the shape selective effect and 2.2 Characterization of the Catalysts
stability of the catalyst for this particular reaction. Corre-
lation between the catalyst characteristics and the catalytic SBA-15, HPW/SBA-15, and Cs-HPW/SBA-15 catalysts
behavior has been established in order to demonstrate the with different loadings were characterized using various
effect of its intrinsic properties on the activity. In addition, techniques to study the physical, chemical and morphological
effects of process variables such as catalyst loading, reac- properties of the catalysts. Nitrogen adsorption–desorption
tion temperature and glycerol-to-lauric acid molar ratio have analysis was performed using a Quanta-chrome Autosorb
been investigated to gain better understanding of the glycerol 1C surface analyzer to investigate the surface characteris-
esterification reaction toward the production of monolaurin. tics of all synthesized catalysts. Fourier transform infrared
The spent catalyst has also been studied to investigate the spectroscopy (FTIR) analysis was conducted by means of a
stability of the morphological and chemical properties of the Shimadzu IR Prestige-21 system using the KBr pellet tech-
catalyst after three consecutive runs. nique. Scanning electron microscopy (SEM) images were
taken using a Quanta FEG 450 SEM, and energy-dispersive
X-ray (EDX) analysis was also performed using the same
2 Experimental equipment. Thermal gravimetric analysis (TGA) was carried
out on a PerkinElmer (TAC 7/DX). The analysis was carried
2.1 Synthesis of Cs-HPW/SBA-15 Catalysts out by raising the temperature from 0 to 900 ◦ C at a ramping
rate of 20 ◦ C/min under 50 mL/min of O2 .
The Cs-HPW/SBA-15 catalysts were synthesized via two-
sequential-step post-impregnation and incipient wetness
technique [9]. The parent SBA-15 was prepared following 2.3 Catalytic Activity
a method described in the literature [10]. 4 g of Pluronic
P123 was dissolved in a solution containing 125 ml of 1.9 The esterification reaction was carried out in a three-necked
M HCl. The solution was stirred at 40 ◦ C for 2 h, and 8 g flask reactor equipped with a magnetic stirrer bar, a thermo-
of tetraethyl orthosilicate (TEOS) was slowly added into the couple, a sampling neck and a tube connected to a vacuum
mixture under a vigorous stirring at 40 ◦ C for 22 h. The solu- pump. In a typical procedure, lauric acid, glycerol and spec-
tion was then transferred into a Teflon bottle and left for ified amount of catalyst were added into the reactor and
another 24 h at 100 ◦ C. The solid product was then filtered, heated to the desired temperature under a reduced pressure
washed with deionized water and dried at room temperature of 50.8 cmHg. Then, the mixture was stirred at 750 rpm for
for 24 h which was then followed by calcination in air at up to 6 h. The experimental conditions employed in this
500 ◦ C for 6 h. For the preparation of 10 wt%Cs-HPW/SBA- work are summarized in Table 1. After the reaction, the
15 catalyst, 0.0304 g of CsNO3 was dissolved in 10 mL of product samples were analyzed using a gas chromatograph
water. Then, 1 ml of the solution was added every 2 min to the (Agilent Technologies 7890A GC system) equipped with a
SBA-15 under a constant stirring. The formed paste was then CP-Sil 5CB (15 m × 0.32 mm × 0.1 mm) column. The lauric
stirred for 20 min until the free-flowing powder was formed. acid conversion and monolaurin selectivity were calculated
The resulted material was dried at 100 ◦ C for 4 h to remove the according to the method proposed by Hermida et. al [2].
remaining solvents and subjected to calcination at 300 ◦ C for
4 h. The obtained product was then impregnated using 0.1795
g of 12-tungstophosphoric acid (HPW) in 10 ml of ethanol. Table 1 Summary of experimental conditions
The resulted product was dried in an oven at 100 ◦ C for 4 h and
Variables Value
then calcined at 300 ◦ C for another 4 h. For a variation of Cs-
HPW concentration (20, 30 and 40 wt%), the catalysts were Catalyst loadings (wt%) 10, 20, 30, 40
synthesized as discussed earlier with the appropriate amount Catalyst concentration with respect to reactants (wt%) 1.5, 2.5, 3.5
of Cs-HPW and their corresponding loadings are denoted Operating temperature (◦ C) 150, 160, 170
as xCs-HPW/SBA-15; x stands for the wt% of Cs-HPW. Glycerol/fatty acid molar ratio 2:1, 3:1, 4:1
For comparison, HPW/SBA-15 catalyst was synthesized via

123
Arabian Journal for Science and Engineering (2018) 43:5771–5783 5773

3 Results and Discussion deposition of active sites inside the pore could have reached
its limit and as a result, the deposition mainly occurred on
3.1 Catalyst Characterization the outside of the pore structure. Nevertheless, this result
indicated that the Cs-HPW was successfully introduced into
3.1.1 Surface Analysis the internal pores of the SBA-15. Meanwhile, all catalysts
exhibited large pore sizes within the range of 58–61 Å.
The surface analysis results of the SBA-15, HPW/SBA-15, Although the difference can be considered as negligible, a
and Cs-HPW/SBA-15 catalysts with different loadings are slight increase in the pore size would give significant effect
represented in Table 2. Based on the results, it can be con- on the catalytic activity in terms of reactant’s conversion and
cluded that the synthesized SBA-15 having a surface area of selectivity, as demonstrated in this work. For the reaction
873 m2 /g and average pore size of 61 Å is a mesoporous mate- involving bulky reactants, pore size plays an important role in
rial. All catalysts had high specific total surface areas in the facilitating the interaction between the molecules with active
range of 383–671 m2 /g. This finding indicated that all cata- sites within the pore. Small pore sizes could cause a geo-
lysts had a large interfacial surface of contact which allow the metrical constraint in accommodating the reactant molecules
interaction between the reactants with the active sites. Never- within the pores while larger pores will minimize diffusion
theless, there was a slight increase in the total surface area of limitations for bulkier molecules such as long-chain fatty
the unmodified HPW/SBA-15 and 10 wt%Cs-HPW/SBA-15 acid.
in which it only increased from 630 to 671 m2 /g. Niiyama The nitrogen adsorption–desorption isotherms of the
et al. [12] reported that the addition of the second group of SBA-15 support and the Cs-HPW/SBA-15 catalysts are
large metal ions like Cs+ (Cs salts of HPAs) on the SBA- shown in Fig. 1. All isotherms exhibited the typical type IV
15 could improve the specific surface area of the catalyst. isotherm (according to the IUPAC classification) [11]. Fur-
However in this study, the increment was minimal so that it thermore, all isotherms exhibited the typical hysteresis loops
can be considered as negligible. This finding suggested that of H1 at relative pressures between 0.4 and 0.9. The hys-
the pore structure of the catalysts remained intact despite the teresis type H1 isotherms with narrow pore size distributions
addition of Cs to indicate the stability of the porous struc- are observed and this finding proved that the catalysts were
ture. porous materials with agglomerated or uniformly packed
Furthermore, the BET surface area for all Cs-HPW/SBA- spheres and regular arrangement. Adsorption and desorption
15 catalysts decreased with increasing Cs-HPW loading. As branches in the hysteresis curve formed are nearly parallel
the Cs-HPW loading on the SBA-15 support was increased with SBA-15 to indicate the formation of the regularly shaped
from 10 to 40 wt%, the BET surface area decreased from pores during the synthesis. Hence, the hysteresis curves in
671 to 383 m2 /g. This phenomenon is in agreement with the multilayer ranges of physisorption, the isotherms were
a previously reported work [13]. Similar trend was also related with capillary condensation in the mesopores [2].
observed for the pore volume in which it decreased from 0.89 Besides, the sharp inflection on the adsorption branch of all
to 0.53 cm3 /g. However, the pore volumes for 20 wt%Cs- the isotherms indicates that the catalysts have relatively high
HPW/SBA-15 and 30 wt%Cs-HPW/SBA-15 catalysts were uniform mesopore sizes distribution.
found to be similar (0.75 cm3 /g). At 30 wt% loading, the Meanwhile, Fig. 2 shows the BJH pore size distribution
of the support and catalysts. All catalysts showed nar-
row ranges of pore size distribution to prove the catalysts
have regular and ordered pore structure [14]. As shown in
Table 2 Surface characteristics of SBA-15, HPW/SBA-15 and
Cs-HPW/SBA-15 catalysts with different Cs-HPW loading Fig. 2, the SBA-15, 20 wt%Cs-HPW/SBA-15 and 40 wt%Cs-
HPW/SBA-15 exhibited smaller peaks at lower pore sizes.
Catalyst Surface area Total pore volume Average pore
(m2 /g) (cm3 /g) size (Å) This was attributed to the tensile strength effect possessed
by the catalysts during the nitrogen adsorption–desorption
SBA-15 873 0.99 61 [15,16]. The tensile strength hypothesis describes that at a
HPW/SBA-15 630 0.84 61 certain pressure, a closure of the hysteresis loop will form
10 wt%Cs- 671 0.89 60 due to the physical properties of the adsorption. It is basi-
HPW/SBA-15 cally because of the “ink-bottle-shape” of the pores and due
20 wt%Cs- 556 0.75 58 to the small size of the pore throats [17]. As the name sug-
HPW/SBA-15
gests, this type of pore opening which was narrower caused
30 wt%Cs- 506 0.75 61
the absorbed N2 to have some difficulty during the desorption
HPW/SBA-15
process. This might be attributed to the uneven silica distri-
40 wt%Cs- 383 0.53 59
HPW/SBA-15 bution and excessive silica at that particular region, forming
silica plug [18].

123
5774 Arabian Journal for Science and Engineering (2018) 43:5771–5783

Fig. 1 Isotherm profiles of SBA-15, HPW/SBA-15 and Cs-HPW/SBA-15 catalysts with different Cs-HPW loading

Fig. 2 BJH pore size distributions of SBA-15, HPW/SBA-15 and Cs-HPW/SBA-15 catalysts with different Cs-HPW loading

3.1.2 Fourier Transform Infrared Spectroscopy (FTIR) [2]. Meanwhile, the unmodified HPW/SBA-15 showed the
spectrum containing the vibrations bands at 1082, 981, 897
Figure 3a shows the FTIR spectra for the SBA-15 support and 803 cm−1 which could be correlated to the stretch-
and the unmodified HPW/SBA-15 while Fig. 3b shows the ing vibrations of P–Oa , W=Od , W–Ob –W and W–Oc –W,
FTIR spectra for all Cs-HPW/SBA-15 with different Cs- respectively [19]. Furthermore, the complex Keggin struc-
HPW loadings. The impregnation of Cs salts of HPW into ture was effectively preserved in all catalysts regardless of
the SBA-15 support evidently affected the structure of the the different concentrations of Cs used in the preparation
catalysts due to partial or complete overlapping of bands of catalysts. However, an overlapping was observed for the
of Cs-HPW supported on SBA-15 matrix. Bands between band corresponded to W–Ob −W at 805–887 cm−1 in Fig. 3a
800 and 1200 cm−1 are observed for all catalyst which can with a strong absorption band of SBA-15 at 882 cm−1 in
be represented by the presence of PW12 O3− 40 Keggin ion Fig. 3b. Basically, at lower loading, more homogeneous dis-

123
Arabian Journal for Science and Engineering (2018) 43:5771–5783 5775

Fig. 3 a FTIR spectra of SBA-15 and HPW/SBA-15 and b FTIR spectra of Cs-HPW/SBA-15 catalysts with different Cs-HPW loading

persion of the active phase inside the pores of the SBA-15 stretching at 805 cm−1 and a bending vibration of Si–
support could be accomplished compared to the higher load- O–Si at 459 cm−1 are observed to suggest the persever-
ings. ance of the mesoporous structure of the SBA-15 support
Bands between 1082 to 462 cm−1 are observed which even at high Cs-HPW loadings. Bands are detected at
could be correlated to the Si-O-Si bond. Some of these 3451, 3447, 3452, 3440 cm−1 for 10, 20, 30 and 40 wt%
bands are preserved in the synthesized catalysts. How- of Cs-HPW loadings, respectively, and these bands are
ever, due to the strong absorption bands of silica from assigned to the Si–OH bonds that formed through the
SBA-15, some of these bands are widened and partially polar interactions between Si–OH water molecules. Other
overlapped. As shown in Fig. 3b, an asymmetric stretch- polar components such as hydrogen bond with the Si–
ing of Si–O–Si at 1250–1083 cm−1 , a symmetric Si–O–Si

123
5776 Arabian Journal for Science and Engineering (2018) 43:5771–5783

OH molecules could also cause the formation of this band 3.1.3 SEM
[20,21].
Interestingly, all synthesized catalysts including SBA-15 Figure 4 shows the morphology of the SBA-15, HPW/SBA-
support show an intense band at wavelengths of 1634– 15, and Cs-HPW/SBA-15 catalysts. As observed in the
1640 cm−1 to represent the adsorbed water molecules. In figure, all catalysts including the SBA-15 support exhibited
general, the intensity of these bands was found to be well-ordered fibrous structures. All catalysts showed similar-
higher for the Cs-HPW/SBA-15 catalysts as compared to ity in terms of the shape and surface morphology as compared
the unmodified catalyst. This could be attributed to the pres- to that of SBA-15 support (Fig. 4a). Hence, it proved that
ence of hydrophilic Si–OH groups in the support material, the catalysts were successfully synthesized and the desired
which could cause rapid adsorption of water molecules into shape, morphology and structure were successfully formed.
the pores [5]. In addition, air humidity could also cause The synthesized SBA-15 is composed of relatively uni-
the phenomenon [22]. In fact, HPW itself is a hydrophilic form rope-like particles which are aggregated into fiber-like
molecule, and hence the SBA-15 supported Cs-HPW cat- macroscopic structures. This morphology resembles the typ-
alysts would definitely have a higher capability to absorb ical SBA-15 extended fiber-like structure [23] to indicate the
moisture or water molecules from the surrounding due to successful synthesis of the SBA-15 mesopores.
the availability of Si–OH bonds. At the wavelength between There was no deposit on the surface of both HPW/SBA-
1083 and 1085 cm−1 , a band is observed for all catalysts 15 (Fig. 4b) and the 20 wt%Cs-HPW/SBA-15 (Fig. 4d),
which could be assigned to [PW11 O39 ]7− which usually and this was due to the flattening of the crystals. However,
presents at pH greater than 2. Upon modification of HPW there are some traces of deposits on the external surface of
by addition of Cs, the protons in HPW are actually replaced 10 wt%Cs-HPW/SBA-15 (Fig. 4c). This could be attributed
by the large cation in the Cs atom. The cation in Cs to the improper mixing of Cs salt with HPW, which in turn
induces electrostatic anion–anion interactions in the cata- reduced the capability of impregnation of the Cs-HPW into
lyst [6]. Thus, the cation in Cs had replaced the proton the pores of the SBA-15 support leaving out tiny crystals on
in HPW and caused a decrease in the stretching frequen- the surface. At Cs-HPW loadings of 30 and 40 wt%, crystals
cies or vibrations of all the bands, particularly at the bands were deposited on the external surface of the SBA-15 support,
representing the oxygen that was shared around the four as shown by the arrow in Fig. 4e and f, respectively. There
edges. is a greater tendency for clusters of crystals and deposits to
form on the external surfaces of the support at high loadings

Fig. 4 SEM images for a SBA-15, b 20 wt%HPW/SBA-15, c 10 wt%Cs-HPW/SBA-15, d 20 wt%Cs-HPW/SBA-15, e 30 wt%Cs-HPW/SBA-15,

f 40 wt%Cs-HPW/SBA-15 samples (magnification 10kX)

123
Arabian Journal for Science and Engineering (2018) 43:5771–5783 5777

(30 and 40 wt%) in post-impregnation method [24]. Higher anions in the mixture of Cs-HPW/SBA-15 catalysts as the
catalyst loading means that more active phases are available amount of anions increased at higher loadings. Higher dis-
to accommodate the pores of the support, resulted in most of persion in the mixture would result in the better permeation
the pores being occupied. This would result in the crystal- of the Cs-HPW anions into the internal pores of the catalyst.
lization of some of the anions (especially the oxidized form Furthermore, the W element also increased with increasing
of HPW) on the external surfaces. This finding also indi- Cs-HPW loadings. This was mainly contributed by the pres-
cated that the HPW anions were not fully impregnated into ence of higher amount of Cs-HPW anions at higher loadings,
the mesopores due to the size restriction of the SBA-15 and which eventually increased the possibility of tungsten triox-
the accumulation of the excess HPW anions on the external ide (WO3 ) formation on the surfaces.
surfaces of the support.
3.1.5 Thermal Gravimetric Analysis (TGA)
3.1.4 EDX Analysis
Figure 5 shows the TGA analysis of the catalysts and
The chemical composition of silicon (Si), oxygen (O), HPW/SBA-15. The profiles show three stages of weight loss
tungsten (W) and caesium (Cs) elements was detected via when it was heated to 900 ◦ C and these regions could be
EDX analysis, and the results are tabulated in Table 3. assigned to the water elimination (region i), decomposition of
In actuality, the catalyst employed in this study consisted the anhydrous samples by constitutive water removal (region
of Cs2.5 H0.5 PW12 O40 , with the enclosure of the anions in ii) and crystallization process of constitutive oxides for the
between the Si layers. However, the phosphorus (P) was not synthesized catalysts (region iii).
detected in all the synthesized catalysts. The element might At the temperature around 29 to 105 ◦ C (region i), a
be fully entrapped in the internal pores rather than on the drastic weight loss was observed for all catalysts includ-
external surface of the catalyst. Moreover, the amount of P ing HPW/SBA-15. This was mainly due to the loss of
in the total catalyst weight could be extremely low to be water by evaporation process accompanied by the crys-
detected via EDX analysis. Thus, the actual amount of Cs- tallization process of the Cs-HPW/SBA-15 catalysts. It
HPW loading on the support could not be estimated due to could also be observed that the weight loss of the cata-
the absence of P. Furthermore, the amount of O was over- lysts increased in a sequence of 40 wt%Cs-HPW/SBA-15
lapped in both HPW anion and SBA-15 support, making it (4.5 wt%), 30 wt%Cs-HPW/SBA-15 (5.0 wt%), 20 wt%Cs-
impossible to distinguish the amount of O present merely HPW/SBA-15 (6.2 wt%), HPW/SBA-15 (9.0 wt%) and
in the Cs-HPW anion. Nevertheless, the difference between 10 wt%Cs-HPW/SBA-15 (12.4 wt%). At lower Cs-HPW
the designated Cs-HPW loadings from the detected one was loadings, the catalysts tend to be more hydrophilic due to the
1.71, 0.19, 1.61, 1.74 and 1.51 wt% for HPW/SBA-15, 10, availability of the vacant sites in between the Si-OH bonds
20, 30 and 40 wt%Cs-HPW/SBA-15, respectively. Thus, it in the SBA-15. Hence, more water molecules were adsorbed
proved that the incorporation of Cs into HPW with SBA-15 and stored during the synthesis process, and upon heating,
was a success since the error detected was sufficiently less the water molecules would be evaporated, resulting in very
than 8%. The missing percentage of Cs-HPW was attributed significant weight losses. Meanwhile, at higher loadings of
to the missing P element. 30 and 40 wt%, more tungsten trioxide (WO3 ) formed on the
As expected, the difference between the designated Cs- external surface of the SBA-15 support and WO3 could have
HPW loadings from the detected results increased at higher
loading. This might be due to the evenly dispersed Cs-HPW

Table 3 Results of EDX analysis on the SBA-15, HPW/SBA-15 and

Cs-HPW/SBA-15 catalysts with different Cs-HPW loading
Catalyst Component

O (wt%) Si (wt%) W (wt%) Cs (wt%)

SBA-15 57.68 42.32 0.00 0.00

HPW/SBA-15 57.78 23.93 18.29 0.00
10 wt%Cs-HPW/SBA-15 54.50 35.69 8.93 0.88
20 wt%Cs-HPW/SBA-15 50.90 30.70 17.04 1.35
30 wt%Cs-HPW/SBA-15 48.83 22.91 25.85 2.41 Fig. 5 TGA profiles of HPW/SBA-15, 10 wt%Cs-HPW/SBA-15,
40 wt%Cs-HPW/SBA-15 37.49 24.03 35.56 2.92 20 wt%Cs-HPW/SBA-15, 30 wt%Cs-HPW/SBA-15 and 40 wt%Cs-
HPW/SBA-15 catalysts

123
5778 Arabian Journal for Science and Engineering (2018) 43:5771–5783

hindered the absorption of water during the synthesis process to that temperature would be sufficient. Hence, it could be
[21]. concluded that the losses in weight of all the synthesized
The second weight loss appears between 180 and 290 ◦ C catalysts beyond this temperature could be attributed by the
(region ii). In this region, the catalysts basically faced the decomposition of the catalysts themselves. In brief, HPW
removal of the remaining water molecules from the Cs- anions of all the synthesized catalysts did not undergo severe
HPW/SBA-15 catalysts. The loss of water molecules actually decomposition. Hence, Cs-HPW/SBA-15 could be consid-
allowed the crystallization of HPW to eventually form HPW ered as a very thermally stable catalyst.
Keggin ions. Traces of organic compounds such as P-123
during the preparation of the SBA-15 were removed in this
region [25]. The third weight loss was observed between 3.2 Catalytic Performance
290 to 590 ◦ C whereby gradual weight losses were encoun-
tered by all the synthesized catalysts. This finding indicated 3.2.1 Effect of Cs-HPW Loading on SBA-15 Support
the removal of the constitutional water that was formed from
acidic protons and oxygen belonging to the Keggin units from Figure 6 shows the lauric acid conversion and monolau-
the internal pores [11]. Dehydroxylation of silanol groups in rin yield profiles of all Cs-HPW/SBA-15 and HPW/SBA-15
the pure HPW usually occurs at a temperature of 485 ◦ C [26]. catalysts. As shown in Fig. 6a, an increase in the Cs-HPW
Since all the synthesized catalysts did not undergo severe loading significantly increased the conversion of lauric acid.
weight losses, it indicated that all synthesized catalysts were After 6 h of reaction, 40 wt% catalyst shows the highest
thermally stable without significant decomposition of acid conversion which was then followed by 30, 20 and 10 wt%Cs-
sites below 900 ◦ C. Still, they were reasonably thermally sta- HPW/SBA-15 catalysts. This basically fits the theory very
ble up to 200 ◦ C and as these catalysts were not designated well that catalysts with higher active acidic sites should have
to be used at a temperature higher than that, the stability up higher activity. These results also suggested that at higher
Cs-HPW loading (30 and 40 wt%), the catalysts might have

Fig. 6 Effects of Cs-HPW loadings on a lauric acid conversion and b monolaurin yield (Reaction temperature = 160 ◦ C, reaction time = 6 h,
catalyst loading = 2.5 wt%, glycerol/lauric acid = 4:1)

Fig. 7 Effects of catalyst loading on a lauric acid conversion and b monolaurin yield for 20 wt%Cs-HPW/SBA-15 (Reaction temperature = 160 ◦ C,
reaction time = 4 h, glycerol/lauric acid = 4:1)

123
Arabian Journal for Science and Engineering (2018) 43:5771–5783 5779

less vacant or available mesopores. The crystals and deposits 3.2.3 Effect of Reaction Temperature
on the catalyst’s external surface might cause difficulty for
the active acid sites to be accessed during the reaction, espe- Effects of reaction temperature (150, 160 and 170 ◦ C) on the
cially after few hours of reaction. This finding is consistent lauric acid conversion and monolaurin yield were investi-
with the results obtained from the SEM and EDX analyses gated and are shown in Fig. 8. This study was conducted
in Fig. 4 and Table 3, respectively. at a glycerol/lauric acid of 4, reaction time of 4 h using
Figure 6b shows the monolaurin yield obtained after 6 h 2.5 wt% of 20 wt%Cs-HPW/SBA-15 catalyst. In this study,
of reaction. It was observed that higher catalyst loading the continuous removal of water from the reaction mixture
(30 and 40 wt%) gave higher yields at the beginning of the successfully shifted the equilibrium toward the glycerides
reaction (0 to 1 h) but after 2 h, 20 wt%Cs-HPW/SBA-15 formation beyond the thermodynamic equilibrium. As a
catalyst showed the highest yield (41.3%). Although the result, higher conversion of lauric acid was achieved. An
40 wt% catalyst might have the highest conversion, but it was increase in the reaction temperature from 150 to 170 ◦ C sig-
not selective toward monolaurin. Therefore, in this study, nificantly increased the conversion from 46.2 to 71.8%. This
20 wt%Cs-HPW/SBA-15 was chosen as the best catalyst finding is consistent with the fact that for an endothermic
to give high monolaurin yield. Furthermore, the equilib- reaction, the equilibrium conversion generally increases with
rium conversion was achieved after 3 h of reaction and the temperature. This is caused by the high molecular veloci-
increment of the yield for the next hours of reaction was ties of the bulky molecules at elevated temperature which
significantly low. Hence, it would be more economical to consequently increases the entropy [27]. At high reaction
shorten the reaction time to 4 h. temperature, the reactant molecules have adequate energy to
overcome the energy barrier resulting in accelerated rate of
3.2.2 Effect of Catalyst Loading reaction. Furthermore, the kinetic energy is higher to result
in the rapid mass transfer of the bulky molecules within the
Effects of catalyst loading (1.5, 2.5 and 3.5 wt%) on the lau- porous channels. As a result, the number of chemisorption on
ric acid conversion and monolaurin yield are shown in Fig. 7. the active sites will be enhanced. Meanwhile at low temper-
As shown in Fig. 7a, an increase in the catalyst loading from ature of 150 ◦ C, low activity was observed at the beginning
1.5 to 3.5 wt% increased the lauric acid conversion from of reaction. At this temperature, the possibility of effective
64 to 66%. This is attributed to the increasing number of collision was low due to the lower energy possessed by the
active acidic sites available to facilitate the reaction. Based reactant molecules [5]. Since both of the kinetic and poten-
on the trend, higher catalyst loading exceeding 3.5 wt% is not tial energies are low at this temperature, the activation energy
favored in this reaction considering the insignificant increase needed for the conversion of lauric acid would be harder to
in the conversion after 4 h of reaction. At certain catalyst overcome. As for the other temperatures (160 and 170 ◦ C),
loading, there will be excess catalyst acidic active sites than the conversion was almost similar after 1 h of reaction.
required. This would result in no further increment in the Figure 8b shows the yield of monolaurin as a function of
conversion even for a prolonged period. Similar finding was reaction time. As shown in the figure, the yield of mono-
also observed using HSO3 SBA-15 catalyst in the reaction laurin obtained after 2 h at 150 ◦ C was the lowest. This
involving lauric acid and glycerol [2]. might be attributed to the formation of other by-products
Meanwhile, the monolaurin yield as a function of reac- like di- and trilaurin. Meanwhile, an increase in the tem-
tion time is shown in Fig. 7b. The highest monolaurin yield perature by 10 ◦ C resulted in a significance increase in the
(31.6%) was obtained using 2.5 wt% of catalyst loading and monolaurin yield. The yield at 170 ◦ C was found to be 44.9%,
for 3.5 wt% catalyst loading, the monolaurin yield slightly which was notably higher than that of 160 ◦ C (31.6%) and
dropped to 30%. Such observation indicated that 2.5 wt% of 150 ◦ C (28.4%). Since the glycerol esterification is a highly
catalyst loading have had the maximum number of acidic endothermic reaction, higher temperature would improve the
active sites required for the esterification reaction. Further extent of esterification. Based on the conversion and yield
increase in the catalyst loading would result in the formation trends, it can be concluded that the reaction temperature of
of bulkier molecules like di- and trilaurin on the external sur- 170 ◦ C was the best temperature in this study.
face of the catalyst that leads to the lower monolaurin yield.
From the results in Fig. 7b, there was actually not much dif- 3.2.4 Effect of Glycerol-to-Lauric Acid Ratio
ference in the yield for all catalysts. The difference between
those catalysts was less than 1%. Hence, the effect of cat- The effect of glycerol-to-lauric acid molar ratio (2:1–4:1) on
alyst loading could actually be considered as insignificant. the lauric acid conversion and monolaurin yield is shown
Still, the best catalyst loading to yield high activity for this in Fig. 9. The result in Fig. 9a shows that the increase in
particular reaction was 2.5 wt% with respect to the limiting the molar ratio increased the reaction rate to consequently
reactant. improve the lauric acid conversion. This could be explained

123
5780 Arabian Journal for Science and Engineering (2018) 43:5771–5783

Fig. 8 Effects of reaction temperature on a conversion b monolaurin yield shown by 20 wt%Cs-HPW/SBA-15 at different reaction temperatures
(Catalyst loading = 2.5 wt%, reaction time = 4 h, glycerol/lauric acid = 4:1)

Fig. 9 Profiles of a conversion b monolaurin yield shown by 20 wt%Cs-HPW/SBA-15 at different reactant ratios (Reaction temperature = 170 ◦ C,
catalyst loading = 2.5 wt%, reaction time = 4 h, glycerol/lauric acid = 4:1)

based on the fact that in an excess of glycerol, high amount reaction. Toward the end of the reaction, the conversions for
of glycerol was available to allow the reaction with lauric the molar ratio of 3 and 4 were almost similar which were
acid molecules to produce mono-, di- and trilaurin. However, around 69–72%.
the increase in the lauric acid conversion was insignificant Figure 9b shows the yield of monolaurin at different molar
when the molar ratio was increased from 3:1 to 4:1 at 4 h ratios. It was observed that molar ratios influenced the yield
of reaction. Although an increase in the glycerol amount is of monolaurin. The yield increased from 30.3 to 44.9% when
favorable to ensure the forward reaction to form products, the molar ratios were increased from 2:1 to 4:1. This was
the effects were found to diminish beyond a certain amount. attributed to the high amount of glycerol available which
This was attributed to the dilution of the limiting reactant increased the possibility for lauric acid to react with glycerol
(lauric acid) which brought an adverse effect on the lauric within the mesoporous channels [2]. Therefore, the earlier
acid conversion [28]. Although the highest reaction rate was attached molecules like monolaurin that diffused out from
observed for molar ratio of 4, the high reactant ratio only the internal pores would not further react with lauric acid to
benefited in the early hours of reaction. This was when the form di- and trilaurin. This consequently increased the con-
reactants were at high concentrations which increased the centration of monolaurin in the reaction mixture. However,
chance for lauric acid to react with glycerol mainly in the no significant change in the yield was observed for the molar
pores of Cs-HPW/SBA-15 catalysts. As the conversion pro- ratio of 4 after 3 h of reaction. As the reaction prolonged, the
ceeded, the glycerol concentration dropped while the lauric yield was found to be maintained at around 44%. This result
acid concentration was way too low to proceed toward the suggests that the selectivity of monolaurin did not increase at
forward reaction. In addition, an excess of glycerol proba- molar ratios higher than 4, but the formation of di- and trilau-
bly covered the active sites in the catalyst which prevented rin increased instead. A similar observation was reported on
the adsorption of lauric acid. Nevertheless, the increase in the esterification of glycerol with lauric acid in the presence
the conversion was only significant at the beginning of the of HSO3 SBA-15 [29] and HPA supported on SBA-15 cat-

123
Arabian Journal for Science and Engineering (2018) 43:5771–5783 5781

Table 4 EDX analysis of the fresh 20 wt%Cs-HPW/SBA-15 and the

spent catalyst
Catalyst Component

O (wt%) Si (wt%) W (wt%) Cs (wt%) S (wt%)

Fresh 50.90 30.70 17.04 1.35 0.00

Spent 64.89 20.10 11.87 1.73 1.18

concluded that the 20 wt%Cs-HPW/SBA-15 catalyst played

a vital role as a heterogeneous catalyst which showed rather
good catalytic activity in several cycles of reactions without
experiencing severe deactivation. Since two-sequential-step
post-impregnation method was employed in this work, the
catalyst might have formed covalent bonding between the
Cs-HPW to the mesoporous silica matrix. Hence, the effect
of pore blockage or leaching would be reduced significantly.
Fig. 10 SEM images of 20 wt%Cs-HPW/SBA-15 catalyst after 3 cycles Apart from being an insoluble salt (Cs salt of HPW on SBA-
of reaction
15 support), the covalent bond between the active sites with
the support was also proven to play a vital role in preventing
alyst [5]. Therefore, it could be concluded that the reactant leaching [30]. Having these two properties, the catalyst used
ratio affected the initial reaction rate but did not have much in this work could be assured to be very stable and almost
effect on the final monoglyceride selectivity. no leaching of any elements from the solid catalyst would
occur; hence, no further testing was actually required.
3.2.5 Stability Study of the Catalyst However, a slight decrease in the total weight of Si and W
elements were observed after three cycles of reaction to sug-
The 20 wt%Cs-HPW/SBA-15 catalyst was tested in the ester- gest that minor deactivation could have occurred. This might
ification reaction to demonstrate the stability of the catalyst be due to the deactivation of the acidic active sites or defects
in terms of the morphological and chemical properties. After in some of the mesopores in the SBA-15. The decrease in
each reaction, the catalyst was recovered and centrifuged to the Si element indicated that the mesoporous structure of the
separate the catalyst from the liquid mixture. The remaining silica was slightly destructed while the decrease in the W
solid was then washed with toluene and acetone for several element suggested that WO3 that was originally present on
times and dried in an oven at 100 ◦ C. Figure 10 shows the the external surfaces of the catalyst was diminished. Inter-
SEM image of the catalyst after 3 consecutive runs. There estingly, the amount of Cs slightly increased from 1.35 to
was no significant difference in the morphology compared 1.73 wt% after a few cycles. This might be due to the Cs
to the fresh catalyst (as shown in Fig. 4d). Furthermore, no elements from the internal pores that diffused out from the
additional crystals or deposits on the catalyst surface were internal surface to the outer surface which in turn increased
observed after multiple runs to indicate the stability of the the amount of available active sites. Hence, the reaction rate
catalyst. would be increased despite the reduced activity of the catalyst
Meanwhile, Table 4 shows the EDX analysis results for after several runs.
both of the fresh and the spent catalyst after 3 consec- Furthermore, this catalyst showed rather good activity
utive cycles. The results indicated that Cs-HPW/SBA-15 with only 2% loss in its activity after three consecutive
catalyst was suitable to be reused in this reaction. The reduc- runs compared to the other reported HPA supported SBA-
tion in the total catalyst loading was around 4.8 wt% which 15 catalysts. It was previously reported that phosphotungstic
was reasonably low after 3 consecutive cycles. This proved acid loaded on pure SBA-15 showed severe HPW leaching
that the catalyst did not undergo very much leaching after in alcohol oxidation reaction with H2 O2 [31]. In another
each cycle. The presence of P and W elements in the prod- study, phosphotungstic acid loaded on hydrophilic ionic liq-
uct mixture could be considered insignificant because the uid modified SBA-15 also displayed the obvious loss of
total reduction in catalyst loading after 3 cycles of reac- activity in further recycles [32]. Besides, 1.2 wt% of sulfur (S)
tion was very low. This brings into a conclusion that the element was also detected after being reused for three cycles.
leaching of the active sites from the support was minimal This could be attributed to the strong chemisorption of the
to cause the homogeneous reaction. Therefore, it could be catalyst’s active sites with the traces of S impurities which

123
5782 Arabian Journal for Science and Engineering (2018) 43:5771–5783

was simultaneously produced in the gaseous phase during the channels and on the external surface of SBA-15 crystals: prepara-
reaction at high temperature. Although this catalyst showed tion and performance as acidic catalysts. Microporous Mesoporous
Mater. 80, 43–55 (2005)
minor S poisoning, the effect was not detrimental to conclude 8. Landau, M.V.; Rao, P.M.; Thomas, S.; Pitchon, V.; Zukerman,
that the 20 wt%Cs-HPW/SBA-15 catalyst was a stable cata- R.; Vradman, L.; Herskowitz, M.: Application of Cs salt of 12-
lyst which could be reused without severe deactivation of the tungstophosphoric acid supported on SBA-15 mesoporous silica
active sites and its catalytic activity. in NOx storage. Top. Catal. 42, 203–207 (2007)
9. Ibrahim, S.M.; El-Shobaky, G.A.: Catalytic efficiency of cesium
and potassium salts of dodecatungstophosphoric acid supported
on silica and comparison with H3PW12O40/SiO2. Kinet. Catal.
4 Conclusions 49, 484–492 (2008)
10. Chiou, J.; Liu, S.; Ho, K.; Huang, H.; Tang, C.; Wang, C.: Ca-
modified Co/SBA-15 catalysts for hydrogen production through
The esterification of glycerol with lauric acid over differ- ethanol steam reforming. Int. Lett. Chem. Phys. Astron. 5, 1–16
ent loading of Cs-HPW/SBA-15 catalysts was successfully (2014)
investigated. The increase in the Cs-HPW loadings increased 11. Gagea, B.C.; Lorgouilloux, Y.; Altintas, Y.; Jacobs, P.A.; Martens,
J.A.: Bifunctional conversion of n-decane over HPW heteropoly
the immobilization of active phase on SBA-15 and improved acid incorporated into SBA-15 during synthesis. J. Catal. 265, 99–
the activity toward high lauric acid conversion and mono- 108 (2009)
laurin yield. Based on the correlation between the catalyst 12. Niiyama, H.; Saito, Y.; Echigoya, E.: In: Proceedings, 7th
characteristics and its activity, it was found that 20 wt%Cs- International Congress on Catalysis, Tokyo, 1980. Kodansha,
Tokyo/Elsevier, Amsterdam (1981)
HPW/SBA-15 catalyst was the best catalyst for this particular 13. Gallegos-Suarez, E.; Pérez-Cadenas, M.; Guerrero-Ruiz, A.;
reaction in which it required shorter reaction time (4 h) to Rodriguez-Ramos, I.; Arcoya, A.: Effect of the functional groups of
obtain 71.8% of lauric acid conversion with 44.9% of mono- carbon on the surface and catalytic properties of Ru/C catalysts for
laurin yield. The highest monolaurin yield was achieved at hydrogenolysis of glycerol. Appl. Surf. Sci. 287, 108–116 (2013)
14. Ertl, G.; Knözinger, H.; Weitkamp, J.: Preparation of solid catalysts.
a temperature of 170 ◦ C using a glycerol-to-lauric acid ratio Wiley, New York (2008)
of 4 and 2.5 wt% of catalyst loading. Furthermore, this cat- 15. Khder, A.E.R.S.; Hassan, H.M.A.; El-Shall, M.S.: Acid catalyzed
alyst was morphologically stable and reusable for up to 3 organic transformations by heteropoly tungstophosphoric acid sup-
cycles during the esterification reaction without any signifi- ported on MCM-41. Appl. Catal. A Gen. 411–412, 77–86 (2012)
16. Olutoye, M.A.; Wong, S.W.; Chin, L.H.; Amani, H.; Asif, M.;
cant deactivation and loss of its activity. Hameed, B.H.: Synthesis of fatty acid methyl esters via the
transesterification of waste cooking oil by methanol with a barium-
Acknowledgements A Research University Grant (814181) and a Dana modified montmorillonite K10 catalyst. Renew. Energy 86, 392–
Inovasi Awal Grant (AUPI00234) from Universiti Sains Malaysia and a 398 (2016)
Transdisciplinary Research Grant Scheme (6762001) from the Ministry 17. Ravikovitch, P.I.; Neimark, A.V.: Experimental confirmation of
of Higher Education Malaysia are gratefully acknowledged. different mechanisms of evaporation from ink-bottle type pores:?
Equilibrium, pore blocking, and cavitation. Langmuir 18, 9830–
9837 (2002)
18. Van Der Voort, P.; Ravikovitch, P.I.; De Jong, K.P.; Benjelloun, M.;
References Van Bavel, E.; Janssen, A.H.; Neimark, A.V.; Weckhuysen, B.M.;
Vansant, E.F.: A new templated ordered structure with combined
1. Yusoff, M.H.M.; Abdullah, A.Z.: Catalytic behavior of sulfated micro- and mesopores and internal silica nanocapsules. J. Phys.
zirconia supported on SBA-15 as catalyst in selective glycerol ester- Chem. B 106, 5873–5877 (2002)
ification with palmitic acid to monopalmitin. J. Taiwan Inst. Chem. 19. Rao-Ginjupalli, S.; Mugawar, S.; Rajan, N.P.; Balla, P.K.; Koman-
Eng. 60, 199–204 (2016) dur, V.R.C.: Vapour phase dehydration of glycerol to acrolein over
2. Hermida, L.; Abdullah, A.Z.; Mohamed, A.R.: Synthesis of mono- tungstated zirconia catalysts. Appl. Surf. Science 309, 153–159
glyceride through glycerol esterification with lauric acid over (2014)
propyl sulfonic acid post-synthesis functionalized SBA-15 meso- 20. Qiao, S.Z.; Bhatia, S.K.; Zhao, X.S.: Prediction of multilayer
porous catalyst. Chem. Eng. J. 174, 668–676 (2011) adsorption and capillary condensation phenomena in cylindrical
3. Alrouh, F.; Karam, A.; Alshaghel, A.; El-Kadri, S.: Direct ester- mesopores. Microporous Mesoporous Mater. 65, 287–298 (2003)
ification of olive-pomace oil using mesoporous silica supported 21. Chen, Y.; Zhang, X.-L.; Chen, X.; Dong, B.-B.; Zheng, X.-C.:
sulfonic acids. Arab. J. Chem. 10(1), S281–S286 (2017) MCM-41 supported 12-tungstophosphoric acid mesoporous mate-
4. Simsek, V.; Degirmenci, L.; Murtezaoglu, K.: Synthesis of a rials: preparation, characterization, and catalytic activities for
silicotungstic acid SBA-15 catalyst for selective monoglyceride benzaldehyde oxidation with H2O2. Solid State Sci. 24, 21–25
production. React. Kinet. Mech. Catal. 117, 773–788 (2016) (2013)
5. Hoo, P.Y.; Abdullah, A.Z.: Direct synthesis of mesoporous 12- 22. Pistonesi, C.; Juan, A.; Irigoyen, B.; Amadeo, N.: Theoretical and
tungstophosphoric acid SBA-15 catalyst for selective esterification experimental study of methane steam reforming reactions over
of glycerol and lauric acid to monolaurate. Chem. Eng. J. 250, 274– nickel catalyst. Appl. Surf. Sci. 253, 4427–4437 (2007)
287 (2014) 23. Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Fredrickson, G.H.;
6. Dias, J.A.; Caliman, E.; Dias, S.C.L.: Effects of cesium ion Chmelka, B.F.; Stucky, G.D.: Triblock copolymer syntheses of
exchange on acidity of 12-tungstophosphoric acid. Microporous mesoporous silica with periodic 50 to 300 Angstrom pores. Sci-
Mesoporous Mater. 76, 221–232 (2004) ence 279, 548–552 (1998)
7. Rao, P.M.; Landau, M.V.; Wolfson, A.; Shapira-Tchelet, A.M.; 24. Liu, S.; Wang, X.; Wang, K.; Lv, R.; Xu, Y.: ZnO/ZnS-PdS
Herskowitz, M.: Cesium salt of a heteropolyacid in nanotubular core/shell nanorods: synthesis, characterization and application for

123
Arabian Journal for Science and Engineering (2018) 43:5771–5783 5783

photocatalytic hydrogen production from a glycerol/water solution. 29. Hermida, L.; Abdullah, A.; Mohamed, A.: Effects of functionaliza-
Appl. Surf. Sci. 283, 732–739 (2013) tion conditions of sulfonic acid grafted SBA-15 on catalytic activity
25. Amani, H.; Ahmad, Z.; Hameed, B.H.: Highly active alumina- in the esterification of glycerol to monoglyceride: a factorial design
supported Cs-Zr mixed oxide catalysts for low-temperature trans- approach. J. Porous. Mater. 19, 835–846 (2012)
esterification of waste cooking oil. Appl. Catal. A Gen. 487, 16–25 30. Abouzari-lotf, E.; Nasef, M.M.; Zakeri, M.; Ahmad, A.; Ripin, A.:
(2014) Composite membranes based on heteropolyacids and their applica-
26. Junior, L.; da Silva, O.; Cavalcanti, R.M.; Matos, TMd; Angelica, tions in fuel cells. In: Inamuddin, D., Mohammad, A., Asiri, A.M.
R.S.; da Rocha Filho, G.N.; Barros, IdCL: Esterification of oleic (eds.) Organic-Inorganic Composite Polymer Electrolyte Mem-
acid using 12-tungstophosphoric supported in flint kaolin of the branes: Preparation, Properties, and Fuel Cell Applications, pp.
Amazonia. Fuel 108, 604–611 (2013) 99–131. Springer, Cham (2017)
27. Dange, P.N.; Rathod, V.K.: Equilibrium and thermodynamic 31. Macierzanka, A.; Szela̧g, H.: Esterification kinetics of glycerol with
parameters for heterogeneous esterification of butyric acid with fatty acids in the presence of zinc carboxylates? Preparation of
methanol under microwave irradiation. Resour. Effic. Technol. 3, modified acylglycerol emulsifiers. Ind. Eng. Chem. Res. 43, 7744–
64–70 (2017) 7753 (2004)
28. Hashemizadeh, I.; Abdullah, A.Z.: Influence of process conditions 32. Palani, A.; Pandurangan, A.: Esterification of acetic acid over
on glycerol esterification catalyzed by tetra-n-butylammonium- mesoporous Al-MCM-41 molecular sieves. J. Mol. Catal. A Chem.
modified montmorillonite catalyst. Online J. Sci. Technol. 2, 47–51 226, 129–134 (2005)
(2012)

123