Vous êtes sur la page 1sur 9

Fuel Processing Technology 148 (2016) 6775

Contents lists available at ScienceDirect

Fuel Processing Technology


journal homepage: www.elsevier.com/locate/fuproc

Research article

Biodiesel production over Ca, Zn, and Al mixed compounds in xed-bed


reactor: Effects of premixing catalyst extrudates with methanol, oil, and
fatty acid methyl esters
Wayu Jindapon a, Prapan Kuchonthara b,c, Chawalit Ngamcharussrivichai b,c,
a
b
c

Program in Petrochemistry, Faculty of Science, Chulalongkorn University, Patumwan, Bangkok 10330, Thailand
Fuels Research Center, Department of Chemical Technology, Faculty of Science, Chulalongkorn University, Patumwan, Bangkok 10330, Thailand
Center of Excellence on Petrochemical and Materials Technology (PETROMAT), Chulalongkorn University, Patumwan, Bangkok 10330, Thailand

a r t i c l e

i n f o

Article history:
Received 18 November 2015
Received in revised form 21 February 2016
Accepted 21 February 2016
Available online 2 March 2016
Keywords:
Biodiesel
Transesterication
Fixed-bed reactor
Calcium oxide
Catalyst extrudate

a b s t r a c t
We studied the production of biodiesel as fatty acid methyl esters (FAMEs) via transesterication of rened
bleached deodorized palm oil with methanol in a xed-bed reactor, using a new heterogeneous base catalyst
in extrudate form. The catalyst extrudates were prepared using a dissolutionprecipitation method from enamel
venus shells (Meretrix meretrix) in the presence of zinc nitrate and alumina as binder precursors. The effects of
addition of hydroxyethylcellulose as a plasticizer and of the extruder type on the extrusion process and physicochemical properties of the extrudates were studied. The as-prepared catalyst extrudate was a mixture of Ca, Zn,
and Al compounds. Calcination at 400 C generated CaO as the major active phase, which was dispersed on the
surface of an alumina support (ZSAH-400). The maximum FAME yield obtained over ZSAH-400 was 65% in a
xed-bed reactor operated at a methanol:oil molar ratio of 30:1, 65 C, and ambient pressure. The effects of
premixing the catalyst extrudates with methanol, vegetable oil, and commercial methyl esters (CMEs) with different fatty acid chain lengths were investigated to improve FAME formation. The ZSAH-400 extrudate premixed
with different CME types gave a stable FAME yield of ~96.5% throughout the operation. The promotion effect provided by premixing with methanol or oil was much less pronounced. A three-phase diagram, thermogravimetry,
and solubility parameters suggest that the CME-rich layer covering the extrudate surface reduces the mass
transfer limitations caused by immiscibility of the reactants.
2016 Elsevier B.V. All rights reserved.

1. Introduction
Biodiesel is a renewable fuel that has several advantages in terms of
exhaust gas cleanliness, non-toxicity, lubricity, biodegradability, and
petroleum-diesel-like fuel properties. Biodiesel, in the form of fatty acid
methyl esters (FAMEs), is commercially produced via transesterication
of triglycerides, which are present in vegetable oils and fats, with methanol, using homogeneous base catalysts. Although the homogeneous
process provides a high reaction rate under mild conditions, it requires
successive steps for catalyst removal and product purication, resulting
in a lower FAME yield and discharge of a large amount of wastewater
[1]. Heterogeneous catalysis is a more promising route for producing
environmentally friendly biodiesel together with the production of
high-purity crude glycerol as a potential biobased building block for
various value-added chemicals.

Corresponding author at: Fuels Research Center, Department of Chemical Technology,


Faculty of Science, Chulalongkorn University, Patumwan, Bangkok 10330, Thailand.
E-mail address: Chawalit.Ng@chula.ac.th (C. Ngamcharussrivichai).

http://dx.doi.org/10.1016/j.fuproc.2016.02.031
0378-3820/ 2016 Elsevier B.V. All rights reserved.

Ca-based catalysts are highly basic and give a high reaction rate in
transesterication under mild conditions [214]. In the past decade, a
large number of Ca-based catalysts consisting of single [29] and mixed
[1014] metal compounds have been investigated for the
transesterication of triglycerides with methanol to produce FAMEs. All
these catalysts are in powder form, and are only appropriate for stirredtank reactors operated under batch conditions. Transesterication in
xed-bed reactors [1520] is of industrial interest because biodiesel and
glycerol can both be continuously produced at a high throughput.
However, powder catalysts cannot be used in xed-bed reactors because
of product contamination with ne particles and a large pressure drop. To
overcome these problems, powder catalysts are formulated as larger
particles with various shapes, e.g., extrudates, pellets, spheres, and
monoliths [21,22].
Ca-based oxides are hard but brittle materials. Their formulation
requires the addition of binders [15,16] and plasticizers [15]. Inorganic
binders (MgO, alumina [15,16], aluminum phosphates [23], clay
(sepiolite) [24], and boehmite [25]) enhance powder contact and the
mechanical strength of the shaped particles obtained after calcination.
Plasticizers (hydroxyethylcellulose (HEC) [15,23], methylcellulose

68

W. Jindapon et al. / Fuel Processing Technology 148 (2016) 6775

[24], and MgO [26]) are used to adjust the viscosity and lubricity of the
catalyst paste to facilitate processing. When a catalyst powder is mixed
with binders and/or plasticizers and then pressed to give a specic
shape, its textural properties and active site content are usually
adversely affected, resulting in a decrease in catalytic performance
[21]. The formulation conditions should therefore be carefully adjusted
to optimize the mechanical, physical, and chemical properties of the
resulting shaped catalysts.
The transesterication of triglycerides with methanol over solid catalysts suffers from serious mass transfer limitations related to immiscibility of the reactants and the number of accessible active sites. Several
techniques have been developed to improve the catalytic performances
of Ca-based oxides used in transesterication [25,15]. CaO has been activated by mixing with methanol prior to the reaction to promote the
formation of methoxide species on its surface [2]. Premixing of CaO
with glycerol in reuxing methanol creates calcium glyceroxides
(Ca(C3H7O3)) as a new catalytically active phase [2]. A small amount
of water added directly to the reaction enhances the FAME yield [2,3].
Tetrahydrofuran [4] and FAME [15] have been added as co-solvents to
improve mass transfer in the reaction system. Premixing CaO with
FAME further increases the reaction rate [5]. The FAME coating on the
catalyst surface prevents deactivation caused by adsorption of atmospheric moisture and CO2 on the active sites [5].
In the present study, we prepared heterogeneous base catalysts in
the form of extrudates via a dissolutionprecipitation method, using
waste seashells as a renewable Ca source. Zinc nitrate and alumina
were used as binder precursors; these were partly converted to spinel
ZnAl2O4 after calcination. The catalyst paste was shaped manually and
with a single-screw extruder. The effects of adding HEC as a plasticizer
on extrusion and the characteristics of the resulting extrudates were
studied. A catalyst extrudate obtained under suitable conditions was
evaluated for its catalytic properties in the transesterication of palm
oil with methanol in a continuous-ow xed-bed reactor at 65 C and
ambient pressure. The effect of premixing the catalyst extrudates with
methanol, vegetable oil or commercial methyl esters (CMEs) with
different fatty acid chain lengths on the transesterication was also
investigated.
2. Experimental
2.1. Preparation of Ca, Zn, and Al mixed compound catalysts in extrudate
form
Enamel venus shells (Meretrix meretrix) were washed, ground, and
sieved to particle sizes of b 10 m, followed by calcination in a mufe
furnace at 800 C for 2 h. Zn(NO3)26H2O (commercial grade, PPM
Chemical) was dissolved in deionized water, and the solution pH was
adjusted to 1 by adding 1 M HNO3 (AR grade, JT Baker). The calcined
seashells were vigorously agitated in the acidied Zn(NO3)2 solution
using an overhead stirrer for 2 h. Al2O3 (commercial grade, PPM
Chemical) was added to the mixture, and stirring was continued for
1 h. The mass ratio of calcined seashell: Zn(NO3)26H2O:Al2O3 was
1:0.6:0.4. HEC (commercial grade, Thai Specialty Chemical Co., Ltd.)
was added to the white slurry. The amount of HEC was varied between
0 and 5 wt.% based on the total solid mass. The resulting mixture was
continuously stirred and heated to remove water until a paste was
formed. The catalyst paste was shaped into continuous rods using a
manual extruder equipped with a 2-mm hole die or a Bonnot singlescrew catalyst extruder equipped with a 3.5-mm hole die. For the
single-screw extruder, the screw speed and frequency were set at
40 rpm and 40 Hz, respectively. The extruded rods were dried at
100 C for 2 h and cut to 5 mm lengths. Finally, the extrudates were
calcined at 400 C for 2 h at a heating rate of 3 C min1 using a mufe
furnace. The catalysts were denoted by ZSAH or ZSAH-400 to represent
the as-prepared extrudate or the extrudate calcined at 400 C,
respectively.

2.2. Catalyst characterization


Elemental analysis of the waste seashells was performed using X-ray
uorescence (XRF) spectroscopy (JEOL ED-2000 energy-dispersive
X-ray uorescence spectrometer). The crystalline structures of the
seashells and the prepared catalysts were investigated using powder
X-ray diffraction (XRD; Rigaku DMAX 2200/Ultima+diffractometer,
Cu K radiation). The XRD patterns were recorded at room temperature
with a 0.02 step size over the 2 range of 1080. The catalyst
extrudates were crushed to ne powders prior to analysis. The patterns
were identied using the Joint Committee on Powder Diffraction Standard (JCPDS) cards. Thermogravimetric/differential thermal analyses
(TG/DTA) of the raw seashells and as-prepared catalysts were performed using a PerkinElmer Diamond thermogravimeter, from room
temperature to 1000 C at a temperature ramping rate of 8 C min1
under a dry N2 ow (10 mL min1).
Optical microscopy (OM) was used to observe the extrudate surface
along the length and cross-sectional area. The snap-shot images were
analyzed using the S-viewer program. The catalyst morphology was
investigated using scanning electron microscopy (SEM; JEOL
JSM-5800LV). The extrudate surface was sputter coated with gold
before SEM analysis. The elemental composition of the extrudate
surface was investigated using energy-dispersive X-ray spectroscopy
(EDX; Shimadzu EDX-720/800HS).
Pulse-chemisorption of CO2 was performed using an AutoChem II
2920 chemisorption analyzer to determine the total basicity of the
catalyst. The samples were pretreated in situ at 400 C for 2 h under an
Ar ow (50 mL min1). CO2 diluted with Ar (10 vol.%) was then injected
into the sample to adsorb CO2, and then excess CO2 was purged by an Ar
ow. The injection and purging were repeated until CO2 adsorption
reached saturation. The amount of basic sites was determined from the
accumulated amount of CO2 that disappeared in each step, measured
using a thermal conductivity detector (TCD). The textural properties of
the catalyst extrudates were determined by N2 physisorption using a
Micromeritics ASAP 2020 surface area and porosity analyzer. The samples
were degassed at 300 C for 2 h prior to the measurements. The specic
surface area (SBET) was calculated using the BrunauerEmmettTeller
(BET) equation and the total pore volume (Vp) was estimated from the
amount of N2 adsorbed at a relative pressure of about 0.990.
The crushing strengths of the catalyst extrudates were tested according to ASTM D-4179-82. The sample (~15 g) was placed in a sample
holder and compressed using a mobile piston at an initial force of 100 N.
The applied force was increased stepwise to 200, 400, and 1000 N until
the extrudates broke, to give a 0.5 wt.% catalyst powder as determined
quantitatively by sieving.
2.3. Miscibility test of methanoloilCME system
Methanol (commercial grade, 99.5%), rened bleached deodorized
(RBD) palm kernel oil (Chumporn Palm Oil Industry PLC), and C12C14
CMEs (Thai Oleochemicals (TOL) Co., Ltd.) were used in the miscibility
test. The fatty acid composition and some basic properties of the RBD
palm kernel oil are summarized in Table S1 in the Supplementary information (SI), while the composition of C12C14 CMEs is shown in Table S2
(SI). All three components were mixed at different volume ratios; the
total volume was maintained at 10 mL. After vigorously stirring at
room temperature for 1 h, the mixture was left for 1 h, and the phase
formed was recorded as a miscible phase, an emulsion, or separate
phases. The results obtained were presented as a three-phase diagram.
2.4. Transesterication procedure
The transesterication of RBD palm oil (Chumporn Palm Oil Industry
PLC) with methanol was performed in a continuous-ow xed-bed reactor. The fatty acid composition and important properties of the RBD
palm oil are shown in Table S1 (SI). A glass column with a 300-mm

W. Jindapon et al. / Fuel Processing Technology 148 (2016) 6775

height and a 30-mm diameter was used as a reactor in which 135 mL of


the catalyst extrudate was packed. The top and bottom of the catalyst
bed were terminated with a layer of glass beads (average diameter
3 mm). The reaction temperature was set at 65 1 C, using a heating
band equipped with an insulator jacket and a thermocouple, and connected to a temperature controller. During the start-up period, the
reactor column was lled with methanol, and was drained once the temperature stabilized. Fresh methanol and palm oil were separately fed
into the bottom of the column at constant ow rates using peristaltic
pumps (Longer, BQ50-1J). The liquid hourly space velocity (LHSV) was
xed at 1.43 h1 and the methanol:oil molar ratio was kept constant
at 30:1. The efuent overowing from the top of the glass column was
collected in 50 mL screw-cap glass bottles at different time intervals.
Excess methanol was removed from the product mixture using a rotary
evaporator, and the FAME phase was recovered without washing. A simplied diagram of the xed-bed reactor system is shown in Fig. S1 (SI).
The FAME composition was determined by gas chromatography
(GC), using a Shimadzu 14B gas chromatograph equipped with a
30-m DB-Wax capillary column and a ame ionization detector. The
FAME yield was calculated according to the standard method (EN
14103) [8,11] using n-heptane (99.8%, Riedel-deHan) as the solvent
and methyl heptadecanoate (99.5%, Fluka) as a reference standard.
2.5. Premixing of ZSAH-400 extrudate with methanol, palm kernel oil, or
CMEs
Methanol, RBD palm kernel oil, and CMEs with different compositions
(Thai Oleochemicals (TOL) Co., Ltd.) were premixed with the catalyst
extrudate to study the effects of adsorption of these substances by the
extrudate on the transesterication of palm oil with methanol. The compositions of different CMEs are shown in Table S2 (SI). After calcination
and cooling to room temperature in a desiccator, the ZSAH-400 extrudate
was immersed in methanol, palm kernel oil, or CMEs at a solid:liquid mass
ratio of 70:30 and stored at room temperature overnight. The saturated
ZSAH-400 extrudates were then packed into the glass reactor, and
transesterication was performed using the procedure described in
Section 2.4, but without lling the reactor with methanol during the
start-up period. When C8C10 and C12C14 CMEs were used in the
premixing, the yields of FAMEs obtained from the reaction were unambiguously determined because the palm oil consisted mainly of C16 and C18
fatty acids (Table S1: SI). However, for premixing with C16C18 CMEs,
the total amount of FAME determined by GC was directly subtracted by
the amount of CME used prior to calculation of the FAME yield.

69

the plasticity of the catalyst paste, and facilitate extrusion [23]. The mechanical strengths of the extrudates obtained after calcination at 400 C
were signicantly lower than those of the dried extrudates. In particular, the extrudates were destroyed or easily broken at high HEC loadings
(Fig. S2(C) and (D): SI). Ketcong and coworkers found that burning off
the HEC in the extrudates by calcination in air increased the porosity
and reduced the mechanical strength [15]. The extrusion using a
single-screw extruder gave denser and harder extrudates because of
the higher pressure applied to catalyst paste. The high-shear mixing
provided by the screw extruder also breaks down particle agglomerates,
giving smaller particles, each of which is entirely coated with a liquid
layer [22]. The extrudates prepared using a single-screw extruder
showed better shape retention than those obtained using a manual extruder after calcination. The addition of HEC also prevented phase separation between solid particles and liquid media during extrusion with
the single-screw extruder, otherwise heat was generated. Formulation
of the extrudate form of the catalyst was therefore subsequently performed using 3 wt.% HEC and a single-screw extruder.
3.2. Catalyst characterization
XRF spectroscopy showed that the raw seashells consisted of 68.6%
CaO, 0.5% MgO, 0.8% SiO2, and 0.1% Fe2O3. Before calcination, the Ca
compounds present were CaCO3 and Ca(OH)2 (Fig. S3(a): SI). TGA
showed that the amounts of CaCO3 and Ca(OH)2 were 62.1% and
37.8%, respectively (not shown). After calcination of the raw seashells
at 800 C, these Ca phases were completely converted to CaO
(Fig. S3(b): SI). The XRF and XRD results indicate that the seashells calcined at 800 C could be used as a CaO precursor for the preparation of
Ca-based mixed compound catalysts.
The thermal decomposition pattern of ZSAH prepared without adding
HEC is shown in Fig. 1. The corresponding phase composition is summarized in Table 2. Decomposition of the as-prepared ZSAH occurred in multiple steps. The assignment of each step was made by comparing the
decomposition pattern with those of authentic samples and related
data in the literature [9,10,2731]. The major phase was Ca(OH)2
(36.4 wt.%). Some Zn2+ was precipitated as Zn5(OH)8(NO3)22H2O. The
ZnO generated by heating Zn5(OH)8(NO3)22H2O at N 300 C showed
low activity in transesterication [27,29]. The weight loss at 280 C
corresponded to dehydration of Al(OH)3 to Al2O3 [28]. The presence of
CaZn2(OH)62H2O and Ca(NO3)2 in the as-prepared catalyst indicated
that Ca2+ was dissolved from CaO in the calcined seashells.
CaZn2(OH)62H2O was formed via coprecipitation of Ca2+ and Zn2+ at
pH 1113 [31]. It was converted to a mixture of CaO and ZnO at 400 C

3. Results and discussion


3.1. Formulation of catalyst extrudates
The effects of the extruder type and amount of HEC used in the
catalyst formulation on the characteristics of the resulting extrudates
are summarized in Table 1. The catalyst paste could not be shaped
into continuous rods using a manual extruder without HEC addition
(Fig. S2(A) and (B): SI). The addition of HEC was needed to enhance
Table 1
Effects of extruder type and HEC amount on extrudate characteristics.
Extruder

Manual

Single screw

a
b

HEC amount
(wt.%)

Extrusion

0
3
5
0
3
5

Faileda
Easy
Easy
Difcult
Easy
Easy

Extrudate characteristics
Dried at 100 C

Calcined at 400 C

Not determined
Hard
Hard
Dense and hard
Dense and hard
Hard

Not determined
Very brittle
Brokenb
Hard
Hard
Very brittle

See the Supplementary information, Fig. S2(A) and (B).


See the Supplementary information, Fig. S2(C) and (D).

Fig. 1. Weight loss and DTG curves of uncalcined ZSAH prepared without adding HEC.

70

W. Jindapon et al. / Fuel Processing Technology 148 (2016) 6775

Table 2
Phase compositions, identied using TGA, of uncalcined ZSAH and ZSAH-400 prepared
without HEC addition.
Weight loss
temperature
(C)

Uncalcined ZSAH
Phase

Composition Phase
(wt.%)

0100
100160
160220
220340

340450
450590
590760

H2O (moisture)
Zn5(OH)8(NO3)22H2O
CaZn2(OH)62H2O
Al(OH)3
Al2O3 a
Ca(OH)2
Ca(NO3)2
CaCO3
Total

3.5
5.1
7.8
4.8
16.8
36.4
14.0
10.1
98.5

ZSAH-400

ZnO
CaO/ZnO

Al2O3

Composition
(wt.%)

4.1
1.8/5.1
24.7

CaO
34.5
Ca(NO3)2 17.3
CaCO3
12.5
100

Corresponds to the quantity added to the synthesis mixture.

[10]. CaO is a highly active phase for the transesterication of vegetable


oils with reuxed methanol [29], while ZnO is insufciently basic to
catalyze the reaction under this condition [29]. The majority of CaO was
derived from dehydration of Ca(OH)2 at 340450 C. Totally, the CaO content of the calcined extrudate was estimated to be 36.3 wt.%. The remaining Ca(NO3)2 and CaCO3 were inactive in transesterication under
reuxing methanol conditions [8]. The TGA results for ZSAH prepared
with HEC addition are not presented here because decomposition of
HEC at 200400 C interfered with quantication of important weight
losses (Fig. S4: SI).
The XRD patterns of the catalyst extrudates before and after calcination at 400 C are shown in Fig. 2. The metal phases found in the asprepared ZSAH were Ca(OH)2, CaZn2(OH)62H2O, and CaCO3. Due to
their low crystallinity, Zn5(OH)8(NO3)22H2O, Ca(NO3)2, Al(OH)3, and
Al2O3 were not detected by XRD. When the catalyst extrudates were calcined at 400 C, CaZn2(OH)62H2O disappeared concomitantly with the
formation of CaO and ZnO as new phases. CaO was also generated from
decomposition of Ca(OH)2. However, CaO was transformed back to
Ca(OH)2 by hydration with atmospheric moisture during the XRD analysis. ZnAl2O4 can be formed via a solid-state reaction of ZnO and Al2O3 at
high temperatures [32]. The presence of ZnAl2O4 at moderate

calcination temperatures in our case suggested that Zn2+ was deposited


on Al2O3 as a support. The increased intensity of the peak at 2 = 29.4,
corresponding to CaCO3, should be attributed to chemisorption of CO2
onto CaO during catalyst preparation.
OM images of the catalyst extrudates before and after calcination at
400 C are shown in Fig. 3. The side view and cross sections of the asprepared extrudate have rough surfaces. After calcination at 400 C,
the side-view and cross-section surfaces remained similar. Fig. 4
shows SEM images of the catalyst extrudates before and after calcination at 400 C. The as-prepared extrudate consisted of small particles
(average size 0.50 m) randomly packed along the length and on the
cross-sectional area. Secondary pores with an average size of 20 nm
were well dispersed in the extrudate, as a result of water released during the drying step. After calcination, the small particles in both the side
views and cross sections of the extrudate were bound together, reducing the size of visible secondary pores. These results suggest that the
particles were uniformly distributed in the calcined extrudate.
EDX was used to determine the elemental composition on the ZSAH400 surface. The atomic ratio of Ca:Zn:Al on the surface was found to be
24.2:6.1:1, which was richer in Ca and Zn than the bulk composition
(Ca:Zn:Al = 5.4:1:1.6), as calculated from the amount of metal precursors used in the catalyst preparation. These results conrm that the
Ca2+ and Zn2 + dissolved in the synthesis mixture were precipitated
as different types of Ca, Zn, and Al mixed compounds on the Al2O3 support. Microscale fractures were observed on both sides of the ZSAH-400
surface (Fig. 4). This is probably related to HEC decomposition during
calcination, which liberated water and CO2 from the catalyst extrudates.
The physicochemical properties of as-prepared ZSAH and calcined
ZSAH-400 are shown in Table 3. The bulk density of the catalyst
extrudate was not signicantly changed by calcination at 400 C. An increase in the BET surface area from 4.23 to 6.40 m2 g1 was observed,
probably due to porosity arising from small surface fractures generated
from HEC decomposition. However, the crushing strength of ZSAH-400
was much lower than that of as-prepared ZSAH. It should be noted that
calcination at 800 C further decreased the crushing strength of the catalyst extrudate to 0.02 MPa. Although ZSAH-400 has a low crushing
strength, the shape and size of the extrudate were not signicantly
changed after being used in transesterication in the xed-bed reactor
(Fig. S5: SI). The diffusion of triglycerides and methanol into the interior
active sites of the catalyst extrudate could be promoted through these
fractures. CO2-pulse chemisorption indicated that, as expected, the
basicity of the ZSAH-400 extrudate (9.3 mol g1) was lower than
that of ZSAH-400 powder (27.0 mol g1). Preliminary reaction tests
under batch conditions showed that a high FAME yield (97.9 wt.%)
was attained over the ZSAH-400 extrudate, which was comparable to
that obtained from using the powdery ZSAH-400 (99.9 wt.%) since the
extrudates were broken to small pieces by vigorous magnetic stirring.
3.3. Miscibility of methanoloilCME system

Fig. 2. XRD patterns of uncalcined ZSAH (a) and ZSAH-400 (b) prepared with 3 wt.% HEC
using a single-screw extruder (symbols: = CaCO3, = CaO, = ZnO, = Ca(OH)2,
= ZnAl2O4, and = CaZn2(OH)62H2O).

Since the FAME added into the reaction improved the mass transfer
of the reaction system as a co-solvent effect [15], the miscibility of the
methanoloilCME system was investigated in this work. A threephase diagram for methanol, palm kernel oil, and C12C14 CMEs is
shown in Fig. 5. The C12C14 CMEs were miscible with either methanol
or palm oil in all proportions, but mixing only methanol and palm oil together always gave separate phases. The addition of the CMEs at high
loading to a mixture of methanol and oil induced a miscible phase in
the system. This could be because the molecular structure of methyl esters, in which the ester groups are hydrophilic and the long hydrocarbon chains of fatty acids are hydrophobic, enables them to dissolve
well in both triglycerides and methanol. A FAME can behave like a
surfactant [33], which promotes the formation of microemulsions in
triglycerides-in-methanol and methanol-in-triglycerides systems [5].
The solubility of oil in methanol or that of methanol in oil is therefore
enhanced by CME addition.

W. Jindapon et al. / Fuel Processing Technology 148 (2016) 6775

Fig. 3. OM images of uncalcined ZSAH side view (A) and cross section (B), and ZSAH-400 side view (C) and cross section (D).

Fig. 4. SEM images of uncalcined ZSAH side view (A) and cross section (B), and ZSAH-400 side view (C) and cross section (D).

71

72

W. Jindapon et al. / Fuel Processing Technology 148 (2016) 6775

Table 3
Physicochemical properties of catalysts.
Catalyst

Type

Bulk density
(g cm3)

Crushing strengtha
(MPa)

SBETb (m2 g1)

Vpc
(mm3 g1)

Dpd
()

Total basicitye
(mol g1)

FAME yieldf
(wt.%)

ZSAH
ZSAH-400
ZSAH-400

Extrudate
Extrudate
Powder

1.01
0.96

1.15
0.13

4.23
6.40
n.d.

6.5
9.0
n.d.

52.8
42.7
n.d.

n.d.g
9.3
27.0

49.9
97.9
99.9

a
b
c
d
e
f
g

Determined using ASTM D-4179-82.


BET surface area.
Total pore volume.
Average pore size.
Determined using CO2 pulse chemisorption.
Batch reactor. Reaction conditions: methanol:oil molar ratio, 30:1; catalyst amount, 10 wt.%; temperature, 65 C; time, 3 h.
n.d. means not determined.

Fig. 6 shows the FAME yields for transesterication of palm oil with
methanol over the ZSAH-400 extrudates in the xed-bed reactor. The
feed of methanol and palm oil at a molar ratio of 30:1 corresponded to
a mixture composition of 50 vol.%, at which a separate phase appeared.
The efuent was collected and analyzed to determine the FAME yield
after a time-on-stream of 120 min. The FAME yield gradually increased
from 39% to 65% within 240 min, during which the liquid inside the reactor and the efuent were transformed from methanoloil separate
phases to a miscible liquid. This indicates that the FAME generated
from the reaction enhanced the methanoloil solubility, and promoted
contact between the reactant molecules and the catalyst surface.
The ZSAH-400 extrudate premixed with methanol gave an initial
FAME yield of 40.1 wt.%. This is similar to the reaction performed on
the extrudates without premixing, because the reaction in the pristine
case was started up in the presence of methanol. Premixing with methanol enhanced the FAME yield to 96.8% at 360 min. This result should be
related to the promoting effect of methoxide species formed on the surface of the catalyst premixed with methanol overnight. The formation of
surface methoxides over CaO immersed in methanol was conrmed by
solid state 13C NMR spectroscopy [17]. Even at room temperature, the
generation of methoxide on the CaO surface occurs easily [34]. Although
a larger amount of surface methoxides over the premixed catalyst could
be expected, the initial stage of the reaction is essentially controlled by
diffusion of bulky triglyceride molecules to the hydrophilic catalyst

surface through the adsorbed methanol layer. When the operation


was prolonged, the FAME generated reduced the mass transfer limitation of the reaction system, resulting in an enhancement of FAME yield.
When the extrudate was premixed with palm kernel oil, the FAME
yield increased to 67.3% at 120 min, which was higher than the FAME
yield obtained in the case of premixing the extrudate with methanol.
The result suggests that diffusion of methanol to the catalyst surface
through the adsorbed triglyceride layer is relatively fast, compared to
diffusion of triglycerides to the surface of the catalyst premixed with
methanol, as a result of the hydrophilicity of methanol and the catalyst
surface. The enhancement of FAME yield at the early stage was
explained by an increase in the initial concentration of triglyceride
reactants on the surface of catalyst extrudates, which promoted the
formation of FAME and the methanoloil miscibility. Consequently,
the FAME yield was enhanced to 92.6% within 480 min.
The ZSAH-400 extrudate premixed with C12C14 CMEs gave stable
FAME yields of 96.5% throughout the operation (Fig. 6). To account for
the advantages of premixing the catalyst with C12C14 CMEs, the thermal decomposition patterns of ZSAH-400 extrudates premixed with different organic substances were analyzed. As shown in Fig. 7(A),
evaporative loss of methanol adsorbed in the extrudates occurred
below 100 C. The weight losses observed at temperatures N 400 C
were mainly related to inorganic species such as rehydrated Ca(OH)2,
Ca(NO3)2, and CaCO3 remaining in the calcined catalysts, as shown in
Fig. 1. Decomposition of the premixing oil occurred in multiple steps
in the range 200400 C (Fig. 7(B)). The extrudate premixed with
C12C14 CMEs showed a major weight loss at 174 C (Fig. 7(C)). In a

Fig. 5. Ternary diagram of methanol, palm kernel oil, and C12C14 CMEs, based on volume
fraction.

Fig. 6. Dependence of FAME yield on time-on-stream in continuous transesterication of


palm oil with methanol over ZSAH-400 extrudates without premixing () and premixed
with methanol (), palm oil (), and C12C14 CMEs (). Reaction conditions:
methanol:oil molar ratio, 30:1; LHSV, 1.43 h1; temperature, 65 C.

3.4. Transesterication over ZSAH-400 extrudates premixed with methanol,


palm kernel oil, and CMEs

W. Jindapon et al. / Fuel Processing Technology 148 (2016) 6775

73

Fig. 7. Weight loss and DTG curves of ZSAH-400 extrudates premixed with methanol (A), palm kernel oil (B), and C12C14 CMEs (C), and extrudates premixed with C12C14 CMEs followed
by immersion in methanol (D) or palm kernel oil (E).

subsequent experiment, ZSAH-400 extrudates previously mixed overnight with C12C14 CMEs were immersed in methanol or palm kernel
oil. The decomposition patterns of these two samples are shown in
Fig. 7(D) and (E), respectively. The weight losses corresponding to
C12C14 CMEs disappeared concomitantly with the appearance of patterns derived from methanol or palm kernel oil. These results indicate
that the CMEs adsorbed on the ZSAH-400 extrudates could be replaced
by methanol and vegetable oil, possibly because of the high solubilities
of these reactants in methyl esters (Fig. 5). It should be noted that this
replacement mainly occurred at the early stage of the operation since
the CMEs used in premixing disappeared after 120 min. Subsequently,
a large amount of FAME generated from the reaction promoted contact
between the reactant molecules and the catalyst surface by increasing
the miscibility of methanol and oil.
The compatibility of methanol and palm kernel oil with the CME
layers coated on the catalyst extrudates was explained by considering
the solubility parameters () of these organic compounds; these can
be calculated using Eq. (1) [35]:

G
;
M

where T is the absolute temperature and SM is the entropy of mixing. If


the liquid components have similar solubility parameters, HM
becomes small. The resulting negative value of GM indicates that
dissolution occurs spontaneously.
According to Eq. (1), the solubility parameter of C12C14 CMEs (8.31
(cal cm3)1/2) falls between those of methanol (9.47 (cal cm3)1/2) and
palm kernel oil (7.88 (cal cm 3)1/2), conrming that dissolution of
methanol or oil in the CME phase is more spontaneous than the mixing
of methanol and oil. However, the solubility parameter of methanol is
underestimated, because the effect of strong hydrogen bonding
between methanol molecules is not taken into account in Eq. (1). The
contributions of non-polar (dispersion) interactions (d), polar

where is the density (g mL1), M is the molecular weight (g mol1),


and G is the group molar attraction constant ((cal cm3)1/2 mol1) [35].
The solubilities of two liquid components are related to the heat of
mixing (HM), which is determined by the difference between their
solubility parameters, as shown in Eq. (2):
H M V M 1  2 2 v1 v2 ;

where VM is the total volume of the mixture and v is the volume fraction
of component 1 or 2 in the mixture. Based on thermodynamic considerations, the change in Gibbs free energy on mixing (GM) depends on the
magnitude of HM, as follows:
GM H M  TSM ;

Fig. 8. Dependence of FAME yield on time-on-stream in continuous transesterication of


palm oil with methanol over ZSAH-400 extrudates without premixing () and premixed
with C8C10 (), C12C14 (), and C16C18 () CMEs. Reaction conditions: see Fig. 6.

74

W. Jindapon et al. / Fuel Processing Technology 148 (2016) 6775

Scheme 1. Simplied model of diffusion of triglycerides and methanol to surface of catalyst extrudates without premixing (top) and for premixing with CMEs (bottom).

(dipoledipole and dipoleinduced dipole) interactions (p), and hydrogen bonding (h) were included in the total solubility parameter
(t) developed by Hansen (Eq. (4)) [36]. The calculation indicates that
the total solubility parameter of methanol is 14.5 (cal cm3)1/2 [37],
which is 1.7-fold higher than the solubility parameter of C12C14
CMEs. As revealed in the three-phase diagram (Fig. 5), a high miscibility
of methanol and CMEs in all proportions should be attributed to hydrogen bonding between the hydroxyl group of methanol and the ester
group of CMEs.
2t 2d 2p 2h :

The ZSAH-400 extrudates premixed with the C8C10 and C16C18


CMEs were also used as catalysts in the transesterication of palm oil
with methanol under continuous conditions to show the effect of CME
type. As shown in Fig. 8, a FAME yield of 96.5% was achieved in the
early stage of the operation and maintained over the time-on-stream,
regardless of the type of CME used for premixing. The solubility parameters of C8C18 CMEs were in the range 7.968.51 (cal cm3)1/2, and are
summarized in Table S3 (SI). Based on the overall results, it can be concluded that the advantage of premixing the catalyst extrudates with different CMEs in biodiesel production in a xed-bed reactor arose from a
physical effect dominated by the high solubility of methanol and oil in
the CME phase.
A simplied model of the benecial effects of premixing the catalyst
extrudates with CMEs on transesterication are shown in Scheme 1. The
reaction between triglycerides and methanol on the catalyst surface
occurred via the EleyRideal mechanism [6]. Without premixing, the
surface of the fresh catalyst is covered with methanol used during the
start-up period. The methanol molecules are activated on the basic
sites of the catalyst to form methoxide species. Transesterication
with surface methoxides is determined by diffusion of bulky triglyceride
molecules passing through the methanol-rich layer. When the ZSAH400 extrudate is premixed with CMEs, the diffusion of methanol and triglycerides to the catalyst surface is facilitated by the CME-rich layer, as
suggested by the TGA results (Fig. 7). The mass transfer limitations in
the early stage of the xed-bed operation are therefore decreased by
premixing the ZSAH-400 extrudate with different CMEs. The FAME generated from the reaction enhances the methanoloil solubility, and

promotes contact between the reactant molecules and the catalyst surface. As a result, a stable FAME yield (96.5%) is achieved throughout the
operation.
4. Conclusions
The Ca, Zn, and Al mixed compound catalysts in extrudate form were
successfully prepared by dissolutionprecipitation method. Although
the addition of HEC enhanced the plasticity of the catalyst paste and facilitated extrusion, its decomposition generated microscale fractures,
which decreased the crushing strength of the resulting extrudate. However, the mechanical strength, textural properties, and basicity of the
catalyst extrudate were sufcient for biodiesel production via
transesterication of palm oil with methanol in a xed-bed reactor
under mild conditions. Because triglycerides and methanol are immiscible, the reaction on the catalyst extrudate premixed with either methanol or oil suffered from mass transfer problems. Premixing the catalyst
extrudate with different types of CMEs resulted in a stable FAME yield
of ~ 96.5% throughout the operation. This benecial effect arises from
the high solubilities of both reactants in the CME layer covering the catalyst extrudate.
Acknowledgements
The authors are grateful to the Chumporn Palm Oil Industry PLC and
the Thai Oleochemicals (TOL) Co., Ltd. for donating palm oil and CMEs
with different fatty acid chain lengths, respectively. This research was
funded by Chulalongkorn University through the Ratchadapiseksompoch
Endowment Fund (2013), project no. CU-56-912-EN. Financial and technical supports from the Thailand Research Fund (TRF) under the project
no. IRG5780001 (Faculty of Science, Chulalongkorn University), the Center of Excellence on Petrochemical and Materials Technology
(PETROMAT), Chulalongkorn University, and the PTT Public Company
Limited, are also acknowledged.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.fuproc.2016.02.031.

W. Jindapon et al. / Fuel Processing Technology 148 (2016) 6775

References
[1] D.Y.C. Leung, X. Wu, M.K.H. Leung, A review on biodiesel production using catalyzed
transesterication, Appl. Energy 87 (2010) 10831095.
[2] A. Esipovich, S. Danov, A. Belousov, A. Rogozhin, Improving methods of CaO
transesterication activity, J. Mol. Catal. A 395 (2014) 225233.
[3] X. Liu, H. He, Y. Wang, S. Zhu, X. Piao, Transesterication of soybean oil to biodiesel
using CaO as a solid base catalyst, Fuel 87 (2008) 216221.
[4] S. Gryglewicz, Rapeseed oil methyl esters preparation using heterogeneous catalysts, Bioresour. Technol. 70 (1999) 249253.
[5] M.L. Granados, D.M. Alonso, A.C. Alba-Rubio, R. Mariscal, M. Ojeda, P. Brettes,
Transesterication of triglycerides by CaO: increase of the reaction rate by biodiesel
addition, Energy Fuel 23 (2009) 22592263.
[6] M. Kouzu, T. Kasuno, M. Tajika, Y. Sugimoto, S. Yamanaka, J. Hidaka, Calcium oxide
as a solid base catalyst for transesterication of soybean oil and its application to
biodiesel production, Fuel 87 (2008) 27982806.
[7] S.H. Teo, A. Islam, T. Yusaf, Y.H. Tauq-Yap, Transesterication of Nannochloropsis
oculata microalga's oil to biodiesel using calcium methoxide catalyst, Energy 78
(2014) 6371.
[8] C. Ngamcharussrivichai, P. Nunthasanti, S. Tanachai, K. Bunyakiat, Biodiesel production through transesterication over natural calciums, Fuel Process. Technol. 91
(2010) 14091415.
[9] D.M. Alonso, F. Vila, R. Mariscal, M. Ojeda, M.L. Granados, J. Santamara-Gonzlez,
Relevance of the physicochemical properties of CaO catalysts for the methanolysis
of triglycerides to obtain biodiesel, Catal. Today 158 (2010) 114120.
[10] J.M. Rubio-Caballero, J. Santamara-Gonzlez, J. Mrida-Robles, R. Moreno-Tost, A.
Jimnez-Lpez, P. Maireles-Torres, Calcium zincate as precursor of active catalysts
for biodiesel production under mild conditions, Appl. Catal. B 91 (2009) 339346.
[11] C. Ngamcharussrivichai, P. Totarat, K. Bunyakiat, Ca and Zn mixed oxide as a heterogeneous base catalyst for transesterication of palm kernel oil, Appl. Catal. A 341
(2008) 7785.
[12] X. Yu, Z. Wen, H. Li, S.-T. Tu, J. Yan, Transesterication of Pistacia chinensis oil for biodiesel catalyzed by CaOCeO2 mixed oxides, Fuel 90 (2011) 18681874.
[13] S. Limmanee, T. Naree, K. Bunyakiat, C. Ngamcharussrivichai, Mixed oxides of Ca, Mg
and Zn as heterogeneous base catalysts for the synthesis of palm kernel oil methyl
esters, Chem. Eng. J. 225 (2013) 616624.
[14] M.C.G. Albuquerque, D.C.S. Azevedo, C.L. Cavalcante, J. Santamara-Gonzlez, J.M.
Mrida-Robles, R. Moreno-Tost, E. Rodrguez-Castelln, A. Jimnez-Lpez, P.
Maireles-Torres, Transesterication of ethyl butyrate with methanol using MgO/
CaO catalysts, J. Mol. Catal. A 300 (2009) 1924.
[15] A. Ketcong, W. Meechan, T. Naree, I. Seneevong, A. Winitsorn, S. Butnark, C.
Ngamcharussrivichai, Production of fatty acid methyl esters over a limestonederived heterogeneous catalyst in a xed-bed reactor, J. Ind. Eng. Chem. 20
(2014) 16651671.
[16] C. Ngamcharussrivichai, W. Meechan, A. Ketcong, K. Kangwansaichon, S. Butnark,
Preparation of heterogeneous catalysts from limestone for transesterication of
vegetable oilseffects of binder addition, J. Ind. Eng. Chem. 17 (2011) 587595.
[17] M. Kouzu, J.S. Hidaka, Y. Komichi, H. Nakano, M. Yamamoto, A process to
transesterify vegetable oil with methanol in the presence of quick lime bit functioning as solid base catalyst, Fuel 88 (2009) 19831990.
[18] L. Bournay, D. Casanave, B. Delfort, G. Hillion, J.A. Chodorge, New heterogeneous
process for biodiesel production: a way to improve the quality and the value of
the crude glycerin produced by biodiesel plants, Catal. Today 106 (2005) 190192.

75

[19] Y.H. Chen, Y.H. Huang, R.H. Lin, N.C. Shang, C.Y. Chang, C.C. Chang, P.C. Chiang, C.Y.
Hu, Biodiesel production in a rotating packed bed using K/-Al2O3 solid catalyst, J.
Taiwan Inst. Chem. Eng. 42 (2011) 937944.
[20] M. Di Serio, S. Mallardo, G. Carotenuto, R. Tesser, E. Santacesaria, Mg/Al hydrotalcite
catalyst for biodiesel production in continuous packed bed reactors, Catal. Today
195 (2012) 5458.
[21] J. Benbow, J. Bridgwater, Paste Flow and Extrusion, Oxford University Press, Inc.,
New York, 1993.
[22] C.H. Bartholomew, R.J. Farrauto, Fundamentals of Industrial Catalytic Processes, John
Wiley & Sons, New Jersey, 2006.
[23] J. Freiding, F.C. Patcas, B. Kraushaar-Czarnetzki, Extrusion of zeolites: properties of
catalysts with a novel aluminium phosphate sintermatrix, Appl. Catal. A 328
(2007) 210218.
[24] D.P. Serrano, R. Sanz, P. Pizarro, I. Moreno, P. de Frutos, S. Blzquez, Preparation of
extruded catalysts based on TS-1 zeolite for their application in propylene epoxidation, Catal. Today 143 (2009) 151157.
[25] S. Mller, Extrusion of Cu/ZnO catalysts for the single-stage gas-phase processing of
dimethyl maleate to tetrahydrofuran, J. Catal. 218 (2003) 419426.
[26] B. Kerkwijk, M. Garca, W.E. van Zyl, L. Winnubst, E.J. Mulder, D.J. Schipper, H.
Verweij, Friction behaviour of solid oxide lubricants as second phase in -Al2O3
and stabilised ZrO2 composites, Wear 256 (2004) 182189.
[27] T. Biswick, W. Jones, A. Pacua, E. Serwicka, J. Podobinski, The role of anhydrous zinc
nitrate in the thermal decomposition of the zinc hydroxy nitrates
Zn5(OH)8(NO3)22H2O and ZnOHNO3H2O, J. Solid State Chem. 180 (2007)
11711179.
[28] A.D.V. Souza, C.C. Arruda, L. Fernandes, M.L.P. Antunes, P.K. Kiyohara, R. Salomo,
Characterization of aluminum hydroxide (Al(OH)3) for use as a porogenic agent
in castable ceramics, J. Eur. Ceram. Soc. 35 (2015) 803812.
[29] A. Ziba, A. Pacua, E.M. Serwicka, A. Drelinkiewicz, Transesterication of triglycerides with methanol over thermally treated Zn5(OH)8(NO3)2 2H2O salt, Fuel 89
(2010) 19611972.
[30] C. Cordeiro, G. Arizaga, L. Ramos, F. Wypych, A new zinc hydroxide nitrate heterogeneous catalyst for the esterication of free fatty acids and the transesterication of
vegetable oils, Catal. Commun. 9 (2008) 21402143.
[31] F. Ziegler, C.A. Johnson, The solubility of calcium zincate (CaZn2(OH)62H2O), Cem.
Concr. Res. 31 (2001) 13271332.
[32] N. Vanderlaag, M. Snel, P. Magusin, G. Dewith, Structural, elastic, thermophysical
and dielectric properties of zinc aluminate (ZnAl2O4), J. Eur. Ceram. Soc. 24 (2004)
24172424.
[33] I. Johansson, M. Svensson, Surfactants based on fatty acids and other natural
hydrophobes, Curr. Opin. Colloid Interface Sci. 6 (2001) 178188.
[34] M. Kouzu, J. Hidaka, Transesterication of vegetable oil into biodiesel catalyzed by
CaO: a review, Fuel 93 (2012) 112.
[35] P.A. Small, Some factors affecting the solubility of polymers, J. Appl. Chem. 3 (1953)
7180.
[36] C.M. Hansen, Hansen Solubility Parameters: A User's Handbook, second ed. CRC
Press, Florida, 2007.
[37] L.H. Sperling, Introduction to Physical Polymer science, fourth ed. WileyInterscience, New York, 1986.

Vous aimerez peut-être aussi