Energy Conversion and Management xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Energy Conversion and Management

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

Evaluation of palm oil mill fly ash supported calcium oxide

as a heterogeneous base catalyst in biodiesel synthesis
from crude palm oil
Wilson Wei Sheng Ho a, Hoon Kiat Ng a, Suyin Gan b,⇑, Sang Huey Tan a
a
Department of Mechanical, Materials and Manufacturing Engineering, The University of Nottingham Malaysia Campus, Jalan Broga, 43500 Semenyih, Selangor Darul Ehsan, Malaysia
b
Department of Chemical and Environmental Engineering, The University of Nottingham Malaysia Campus, Jalan Broga, 43500 Semenyih, Selangor Darul Ehsan, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history: A palm oil mill fly ash supported calcium oxide (CaO) catalyst was developed to be used as a heteroge-
Available online xxxx neous base catalyst in biodiesel synthesis from crude palm oil (CPO). The catalyst preparation procedure
was optimised in terms of final calcination temperature and duration. The optimum catalyst preparation
Keywords: conditions were determined as final calcination at 850 °C for 2 h with 45 wt.% loading of calcined calcium
Crude palm oil carbonate (CaCO3). A maximum biodiesel yield of 75.73% was achieved for this catalyst under fixed
Heterogeneous catalyst transesterification conditions. Characterisation tests showed that the catalyst had higher surface area
Transesterification
and basic sites which favoured transesterification. The effects of catalyst loading, methanol to oil molar
Biodiesel
Catalyst reusability
ratio, reaction temperature and reaction time on biodiesel yield and fatty acid methyl ester (FAME) con-
version were also investigated. It was determined that transesterification conditions of 6 wt.% catalyst
loading, 12:1 methanol to oil molar ratio, 45 °C reaction temperature, 3 h reaction time and 700 rpm stir-
ring speed resulted in biodiesel yield and FAME conversion of 79.76% and 97.09%, respectively. Experi-
mental kinetic data obtained from the heterogeneous transesterification reactions fitted the pseudo-
first order kinetic model. The activation energy (Ea) of the reaction was calculated to be 42.56 kJ mol1.
Key physicochemical properties of the produced biodiesel were measured and found to be within the
limits set by EN 14214. The developed catalyst could feasibly be used up to three consecutive cycles after
regeneration using methanol washing followed by recalcination at 850 °C for 2 h.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction using heterogeneous catalysed transesterification include less

Heterogeneous catalysed transesterification has recently been
proposed as an alternative process to homogeneous catalysed
transesterification for biodiesel production. The advantages of
Abbreviations: Al, aluminium; Al2O3, aluminium oxide; ASTM, American Society
for Testing and Materials; BET, Brunauer–Emmett–Teller; Ca, calcium; CaCO3,
calcium carbonate; CaO, calcium oxide; Ca(OH)2, calcium hydroxide; Ca2SiO4,
dicalcium disilicate; CN, cetane number; CO2, carbon dioxide; CPO, crude palm oil;
DG, diglyceride; DOBI, bleachability index; Ea, activation energy; EDS, energy
dispersive X-ray spectroscopy; FAME, fatty acid methyl ester; FFA, free fatty acid;
FID, flame ionization detector; GC, gas chromatography; HHV, high heating value;
H2O, water; ICP-AES, inductively coupled plasma-atomic emission spectroscopy;
JCO, Jatropha curcas oil; K, potassium; k, reaction rate constant; ko, pre-exponential
factor; Li/CaO, lithium doped calcium oxide; MG, monoglyceride; Mg, magnesium;
N2, nitrogen; O, oxygen; P, phosphorus; R, gas constant; R2, correlation coefficient; r,
reaction rate; SEM, scanning electron microscopy; Si, silicon; SiO2, silicon oxide; T,
temperature; t, time; TG, triglycerides; TGA, thermogravimetric analysis; XME,
conversion of methyl ester; XRD, X-ray diffraction; ZnO, zinc oxide.
⇑ Corresponding author. Tel.: +60 3 89248162; fax: +60 3 89248017.
E-mail address: suyin.gan@nottingham.edu.my (S. Gan).
wastewater generation and the ability to recycle or reuse the
catalyst [1]. These advantages could potentially lead to lower pro-
duction costs. Nonetheless, there are two key limitations in the use
of heterogeneous catalysts. Firstly, heterogeneous catalysts have
lower reaction rates as compared to homogeneous catalysts. Sec-
ondly, the catalyst surface active sites are rapidly poisoned upon
exposure to ambient conditions because of chemisorption of car-
bon dioxide (CO2) and water (H2O) onto the surface sites [2]. Thus,
further development of heterogeneous catalysts should focus on
enhancing not only catalytic performance, but also its reusability.
Calcium oxide (CaO) has recently been reported as an active het-
erogeneous base transesterification catalyst due to its high basicity,
low solubility and easy handling [3–6]. The use of CaO can be cate-
gorised into five main groups. The groups are neat CaO, doped CaO,
mixed CaO, waste CaO and supported CaO [4]. With regard to neat
CaO applications, Kouzu et al. [7] has examined the reactivity of
CaO, calcium hydroxide (Ca(OH)2) and calcium carbonate (CaCO3)
in converting soybean oil into biodiesel. Fatty acid methyl ester

http://dx.doi.org/10.1016/j.enconman.2014.03.061
0196-8904/Ó 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Ho WWS et al. Evaluation of palm oil mill fly ash supported calcium oxide as a heterogeneous base catalyst in biodiesel
synthesis from crude palm oil. Energy Convers Manage (2014), http://dx.doi.org/10.1016/j.enconman.2014.03.061
2 W.W.S. Ho et al. / Energy Conversion and Management xxx (2014) xxx–xxx
(FAME) conversions of 93%, 12% and 0% were reported for CaO,
Ca(OH)2 and CaCO3, respectively. However, it was reported that
the catalytic activity of CaCO3 could be improved by employing
thermal activation treatment such as calcination in order to remove
the surface carbonates [8]. To enhance the performance of CaO,
some researchers have tried doping an active ingredient onto CaO
[9,10]. For instance, lithium doped CaO (Li/CaO) was examined by
Alonso et al. [11] in the transesterification of sunflower oil. The
authors indicated that activation of the catalyst at high temperature
was essential. However, when the activation temperature exceeded
500 °C, lixiviation occurred which affected the catalytic perfor-
mance. CaO has also been mixed with other oxides for transesteri-
fication. It has been reported that CaO mixed with zinc oxide (ZnO)
used in the transesterification of palm kernel oil to biodiesel re-
sulted in a high FAME conversion of 94% [12]. In another mixed
CaO attempt, Zhu et al. [13] treated CaO with ammonium carbonate
solution and calcined it at 900 °C to form a solid super base, which
showed high catalytic activity in Jatropha curcas oil (JCO) transeste-
rification. Under the optimum conditions, the conversion of JCO
reached 93%. In order to develop more cost effective CaO based cat-
alysts, waste CaO and supported CaO are preferred. Waste mud
crab, cockle and chicken eggs shells were reportedly utilised as
CaO based catalysts in palm olein transesterification [14–16].
Previously, CaO supported on different palm oil mill waste boiler
ashes were developed and characterised by the authors for the
transesterification of CPO [17]. It was found that the use of CaO sup-
ported on waste fly ash resulted in the best FAME conversion of
94.5% under fixed transesterification conditions [17]. The present
study focuses on further developing and evaluating the waste fly
ash supported CaO as a catalyst to convert CPO to biodiesel. Four as-
pects are covered in this study, all of which have not been reported
in the literature before. Firstly, the wet impregnation method used
in catalyst development is optimised in terms of calcination tem-
perature and duration after metal loading. The motivation for this
optimisation is to increase the catalytic performance and poten-
tially reduce the energy consumption in catalyst preparation. Sec-
ondly, parametric transesterification experimental tests are
carried out to elucidate the effects of catalyst loading, methanol
to oil molar ratio, reaction time and temperature on FAME conver-
sion and biodiesel yield. For this study, both FAME conversion and
biodiesel yield are measured to evaluate the reaction process. In the
majority of the reported papers on transesterification, the effective-
ness of the reaction is solely judged by the FAME conversion
[2,8,14,15,18–20]. Although biodiesel yield is important as an indi-
cator of the process viability, it is seldom measured. A minority of
studies have measured the biodiesel yield [13,21] or both FAME
conversion and yield [22,23] in order to evaluate the transesterifi-
cation reaction. The focus of the parametric tests is to determine
the conditions which would result in the highest possible biodiesel
yield without compromising the FAME conversion which must ex-
ceed 96.5% as specified in the European Standard EN 14214 [24].
Under optimised conditions, the key biodiesel properties are also
determined. Thirdly, the transesterification kinetic using the waste
fly ash supported CaO catalyst is studied in order to determine the
reaction rate constant (k) and activation energy (Ea). Lastly, the
reusability of this heterogeneous base catalyst is examined in light
of the need for development of enhanced reusability of heteroge-
neous catalysts for biodiesel synthesis.
2. Materials and methods
2.1. Materials and chemicals
CPO was provided by Havys Oil Mill Sdn. Bhd., Malaysia. The
accompanying sample analysis report stated that the CPO had a
free fatty acid (FFA) content of 3.7%, a deterioration of bleachability
Please cite this article in press as: Ho WWS et al. Evaluation of palm oil mill fly ash supported calcium oxide as a heterogeneous base catalyst in biodiesel
synthesis from crude palm oil. Energy Convers Manage (2014), http://dx.doi.org/10.1016/j.enconman.2014.03.061
index (DOBI) value of 2.6% and a moisture content of 0.2%. Palm oil
mill fly ash was collected from Seri Ulu Langat Palm Oil Mill Sdn.
Bhd., Malaysia to be used as catalyst support. Methanol (>99% pur-
ity) and CaCO3 (99.95% purity) were purchased from Merck and
Sigma Aldrich, respectively. Meanwhile, an internal standard
methyl heptadecanoate was purchased from Sigma Aldrich.
2.2. Catalyst synthesis
The wet impregnation method was utilised to prepare the cat-
alyst at a fixed composition of 45 wt.% calcined CaCO3 loaded onto
palm oil mill fly ash [17]. In brief, the fly ash was first sieved and
dried at 105 °C for 24 h to eradicate all moisture. Prior to use,
CaCO3 was calcined at 800 °C for 90 min to remove CO2 and gener-
ate CaO. Subsequently, water was added to obtain the active salt
precursor Ca(OH)2. The purpose of dissolving CaO into aqueous
Ca(OH)2 was to create a suitable medium for both the CaO and
fly ash to combine well during mixing. Typically, to obtain
45 wt.% calcined CaCO3 loaded catalyst, 13.5 g of calcined CaCO3
was added to 200 mL of pure deionised H2O in a 3-neck flask under
magnetic stirring at 700 rpm to prepare an aqueous solution of
Ca(OH)2. The solution was heated in a closed system at 70 °C to
minimise water loss. 16.5 g of fly ash was then slowly added to
the solution and mixed vigorously for another 4 h until a homoge-
nous mixture was obtained. The sample was then aged for 18 h to
ensure that the Ca(OH)2 precipitated on the fly ash carrier. The re-
moval of adhering H2O in the mixture was accomplished using a
hot air oven at 105 °C for 24 h. A hot filtration test following the
methodology given by Lempers and Sheldon [25] was carried out
to confirm that the developed catalyst was heterogeneous. Briefly,
the catalyst was filtered from the transesterification reaction mix-
ture halfway through the reaction at 2 h and 80 °C (6 wt.% catalyst
loading and 12:1 methanol to oil molar ratio) and the reaction was
allowed to continue in the absence of the catalyst. The contribution
of homogeneity was ruled out since the reaction came to a halt.
To develop a catalyst with increased catalytic activity and
potentially reduce the energy consumption in catalyst preparation,
the wet impregnation method used was optimised in terms of cal-
cination temperature and duration after metal loading. The dried
catalyst was calcined at temperatures of 850, 900, 950 and
1000 °C in a muffle furnace for two different durations of 2 and 5 h.
2.3. Catalyst characterisation
The basic strength of the catalysts was determined by using the
following Hammett indicators: phenolphthalein (H_ = 9.3), 2,4-
dinitroaniline (H _ = 15) and 4-nitroaniline (H_ = 18.4). Typically,
25 mg of catalyst was mixed with 1 mL of Hammett indicator di-
luted in 20 mL of methanol and equilibrated for 2 h. A colour
change indicated that the catalyst is stronger than the indicator
and vice versa [14]. The synthesised catalysts were examined by
thermogravimetry analysis (TGA) using a Mettler Toledo TGA/
DSC 1 system operating from ambient temperature to 900 °C at a
scanning rate of 10 °C min1 under a nitrogen (N2) flow rate of
15 mL min1. The surface morphology of the developed catalysts
was analysed using scanning electron microscopy (SEM) equipped
with energy dispersive X-ray spectroscopy (EDS). The catalysts
were photographed using SEM in a FEI Quanta 400F FESEM system
operating at 15 kV to estimate the particle sizes and shapes. The
EDS elemental composition (particularly the calcium (Ca) concen-
tration) was semi-quantitatively expressed in terms of weight per-
centages using Oxford-instruments INCA 400. EDS was used due to
the unavailability of more accurate quantification techniques such
as inductively coupled plasma-atomic emission spectroscopy (ICP-
AES). In addition, X-ray diffraction (XRD) was utilised as a quanti-
tative determination of the CaO concentration within the catalysts
 W.W.S. Ho et al. / Energy Conversion and Management xxx (2014) xxx–xxx
[26]. The crystalline structure of the catalysts was also character-
ised via XRD in a PANanalytical X’pert Pro system using Cu Ka radi-
ation (40 mA and 45 kV) equipped with a Xe proportional detector.
The analysis was performed at a scanning rate of 1° min1 (step
size of 0.02°) in the 2h range from 10° to 80°. The specific surface
areas of the fresh and used catalysts were measured using the Bru-
nauer–Emmett–Teller (BET) surface area analysis. The catalysts
were tested for their surface areas and porosities using a Microm-
eritics’ ASAP 2020 porosimetry analyser. This was accomplished
using N2 adsorption–desorption isotherms at liquid N2 tempera-
ture and relative pressures (P/Po) ranging from 0.07 to 0.3 whereby
a linear relationship was maintained. Prior to each measurement,
the sample was degassed at 350 °C for 3 h.
2.4. Catalytic activity evaluation of palm oil mill fly ash supported CaO
catalyst
The palm oil mill fly ash supported CaO catalysts prepared un-
der different final calcination conditions were evaluated in terms
of their catalytic activity in the transesterification of CPO. Fixed
transesterification conditions of methanol to oil molar ratio of
12:1, temperature of 60 °C and stirring speed of 700 rpm for 3 h
were chosen [17]. The catalyst loading was varied from 2 to
8 wt.% and the biodiesel yield was measured.
A single factor parametric study was then carried out to inves-
tigate the effects of catalyst loading, methanol to oil molar ratio,
reaction time and reaction temperature on biodiesel yield and
FAME conversion using the catalyst prepared under optimum con-
ditions. Table 1 shows the nominal parametric settings for all the
transesterification tests along with the variation in each of the
parameters for the single factor parametric study. All experiments
were repeated thrice.
The transesterification experiments were carried out in a
250 mL three-neck round bottom flask equipped with a reflux con-
denser. A thermometer was fitted onto one side-neck while the
other side-neck was used for sampling. The reactor was placed in
a heating mantle with controlled heating and stirring. For all tests,
100 mL (87 g) of CPO was used. The CPO was first heated to 5 °C
above the desired reaction temperature. The desired amount of
fly ash supported CaO catalyst was added to the methanol and
was premixed at 40 °C and 500 rpm for 15 min. The catalyst-meth-
anol mixture was then introduced into the reactor. Stirring was ini-
tiated and the reaction was carried out at the desired temperature
and time.
After the reaction, the solid catalyst was filtered out from the
mixture with the aid of a vacuum pump. The filtered sample was
separated by centrifugation at 4000 rpm for 15 min. Two phases
would be formed, namely an ester-rich phase (top layer) and a
glycerol-rich phase (bottom layer). The glycerol was removed
using a separating funnel while the excess methanol in the ester-
rich phase was removed via a rotary evaporator. The sample from
the ester-rich phase underwent diluted acetic acid washing (90 °C)
to remove all dissolved residues and hot air oven drying (65 °C) to
remove all H2O content. The mass of the final product, i.e. biodiesel
was measured for the determination of yield according to Eq. (1).
Table 1
The nominal and tested parametric settings for the transesterification study.
Reaction parameters Tested settings
Catalyst loading (wt.%) 2, 3, 4, 5, (6)a, 7, 8
Methanol to oil molar ratio (mol/mol) 6:1, 9:1, (12:1)a, 15:1, 18:1
Reaction time (h) (3)a, 4, 5, 6, 7
Reaction temperature (°C) 45, 50, 55, (60)a, 65
a
The settings in parentheses are the nominal settings.
Please cite this article in press as: Ho WWS et al. Evaluation of palm oil mill fly ash supported calcium oxide as a heterogeneous base catalyst in biodiesel
synthesis from crude palm oil. Energy Convers Manage (2014), http://dx.doi.org/10.1016/j.enconman.2014.03.061
3
Weight of biodiesel
Yield ð%Þ ¼ ð1Þ
Weight of oil
2.5. Experimental kinetic study
Transesterification experiments were carried out at different
reaction temperatures (45, 50, 55, 60, 65 °C) under fixed catalyst
loading of 6 wt.% and methanol to oil molar ratio of 12:1. Each
experiment was allowed to run until 6 h to ensure complete con-
version of CPO into FAME. At every 1-h interval, a 2 ml sample
was withdrawn from the reaction mixture for gas chromatography
(GC) analysis. The withdrawn sample was immediately quenched
with 4 ml 0.1 M aqueous acetic acid to neutralise the reaction. It
was then centrifuged at 1600 rpm for 15 min to form a two-phase
mixture comprising an ester phase and an aqueous phase. An
emulsion was formed in between these two phases and its thick-
ness varied proportionally to the amount of diglyceride (DG) and
monoglyceride (MG) formed in the reaction mixture [27]. The
aqueous layer was discarded while the upper ester layer was dried
in an oven at 105 °C for 24 h before GC analysis.
2.6. Gas chromatography (GC) analysis
The FAME content in the biodiesel samples was quantitatively
analysed by GC according to European Standard EN 14103 [28].
The analysis was carried out using a Perkin Elmer Clarus 500 GC
equipped with a flame ionisation detector (FID). The capillary col-
umn used was a highly polar Agilent J&W HP-88 cyanopropyl col-
umn with dimensions of 0.25 mm I.D.  60 m length  0.2 lm film
thickness. Helium was used as the carrier gas. The pressure of the
gas flow was set at 32 psi. The oven temperature was isothermally
held at 185 °C for 15 min. Meanwhile, the injector and detector
temperatures were set at 240 °C and 250 °C, respectively.
50 mg of sample was placed in a 2 mL vial and 1 mL of internal
standard methyl heptadecanoate solution (10 mg/mL) was added
using a pipette. The methyl heptadecanoate solution was prepared
by weighing 500 mg of methyl heptadecanoate in 50 mL of hexane.
1 lL of sample was injected into the GC with a split injection ratio
of 100:1. This quantitative analysis provides verification that the
ester content in biodiesel is greater than 96.5% m/m in accordance
with the specifications reported in EN 14214 [24]. As detailed in
the European standard EN 14103, the ester content (FAME conver-
sion) expressed as a mass fraction in percentage was calculated
using Eq. (2) [28].
  
RA  AIS C IS V IS
Ester content ð%Þ ¼ 100 ð2Þ
AIS m
P
where A = sum of area of all peaks ranging from C14:0 and C24:1;
AIS = internal standard (methyl heptadecanoate) peak area;
CIS = concentration of the internal standard solution, in mg/mL;
VIS = volume of the internal standard solution used, in mL; m = mass
of the sample, in mg.
2.7. Biodiesel physicochemical properties determination
Key physicochemical properties of the biodiesel produced un-
der fixed transesterification conditions of 6 wt.% catalyst loading,
12:1 methanol to oil molar ratio, 45 °C temperature, 3 h time and
700 rpm stirring speed were determined according to American
Society for Testing and Materials (ASTM) standards. The measured
properties included: (i) kinematic viscosity at 40 °C determined
using glass capillary viscometers according to ASTM D 445; (ii)
density measured using the hydrometer method according to
ASTM D 1298; (iii) flash point measured using the rapid equilib-
rium closed cup method according to ASTM D 93; (iv) acid value
4 W.W.S. Ho et al. / Energy Conversion and Management xxx (2014) xxx–xxx
determined by volumetric titration according to ASTM D 664 and;
(v) higher heating value (HHV) estimated using a bomb calorime-
ter accordingly to ASTM D 4868. Due to the unavailability of cetane
number (CN) testing facility on-campus, Eq. (3) which was devel-
oped using Artificial Neural Networks (ANN) by Cheenkachorn
[29] was used to calculate CN:
CN ¼ 33:6 þ 0:539 ðC18 : 0Þ þ 0:303 ðC18
: 1Þ þ 0:0878 ðC18 : 2Þ þ 0:233 ðC22 : 1Þ
where CN indicates the cetane number while (C18:0), (C18:1),
(C18:2), (C22:1) are the mass concentrations of stearic, oleic, lino-
leic and erucic acids, respectively.
2.8. Kinetic modelling
Considering that excess methanol was used in the reaction to
shift the reaction equilibrium towards the product side, the reverse
reaction can be ignored [30]. To simplify the kinetic modelling, the
reaction was assumed to be pseudo-first order reaction and hence
the following rate equations as used by Birla et al. [31] were
applicable:
d½TG
r¼ ¼ k½TG
dt
where r is the transesterification reaction rate (mol mL1 min1), k
is the equilibrium rate constant (min1) and [TG] is concentration of
triglycerides (mol mL1) and t is the time (min).
Integration of Eq. (3) from t = 0 and [TG] = [TG0] till t = t and
[TG] = [TG] gives:
ln½TG0   ln½TGt  ¼ kt
From mass balance,
½TG
X ME ¼ 1 
½TG0 
or,
½TG ¼ ½TG0 ð1  X ME Þ
where XME is the conversion of methyl ester. Eq. (7) can also be ex-
pressed in terms of the following rate equation:
dX ME
¼ kð1  X ME Þ
dt
Integration and rearrangement of Eq. (8) results in:
 lnð1  X ME Þ ¼ kt
By fitting the experimental data to Eq. (9), the rate constants k
at different temperatures can be obtained from the linear slope
through the origin. The activation energy Ea and pre-exponential
factor k0 for the transesterification reaction can be calculated using
the following Arrhenius equation:
Ea
k ¼ k0 eRT
1
where Ea is the activation energy (kJ mol ), R is the gas constant
(8.314 J mol1 K1), T is the temperature (K), and k0 is the pre-expo-
nential factor (min1).
By taking the natural logarithm of Eq. (10), Eq. (11) is obtained:
Ea 1
ln k ¼   þ ln k0
R T
where Ea and k0 can be determined from the slope and intercept,
respectively.
Please cite this article in press as: Ho WWS et al. Evaluation of palm oil mill fly ash supported calcium oxide as a heterogeneous base catalyst in biodiesel
synthesis from crude palm oil. Energy Convers Manage (2014), http://dx.doi.org/10.1016/j.enconman.2014.03.061
2.9. Catalyst reusability
After the first optimised transesterification cycle, the recovered
catalysts were tested with 3 different catalyst regeneration meth-
ods to determine which method provided the best recovery of CaO
and Ca concentrations. The regeneration methods were as follows:
(i) methanol washing; (ii) recalcination without methanol wash-
ing; and (iii) methanol washing followed by recalcination. In the
ð3Þ first method, after completion of reaction, the spent solid catalyst
was separated from the mixture by filtration and washed with
methanol to remove polar compounds on its surface such as glyc-
erol. In the second method, the separated catalyst was directly
recalcined at 850 °C for 2 h without any methanol washing. Lastly,
in the third method, the separated catalyst was washed with meth-
anol first and then dried in the oven (65 °C for 3 h) before recalcin-
ing at 850 °C for 2 h. A total of 5 transesterification cycles were
carried out using the optimum regeneration method to study the
catalyst reusability. The determination of the CaO and Ca concen-
trations of the catalyst after each regeneration method and each
transesterification cycle was carried out using XRD and EDS,
respectively.
3. Results and discussion
ð4Þ
3.1. Characterisation of the catalysts
3.1.1. Basic strength analysis
The catalysts changed the colour of phenolphthalein from col-
ourless to pink, but failed to change the colour of 2,4-dinitroaniline
and 4-nitroaniline. This indicated that the catalysts’ basic strength
ð5Þ was within 9.3 < H_ < 15. As such, the developed catalysts were
considered as a relatively strong base catalyst for transesterifica-
tion although the strength was slightly lesser than the original
CaO (typically 15 < H_ < 18.4 [14]).
ð6Þ
3.1.2. Thermogravimetry Analysis (TGA)
Fig. 1 depicts the TGA profiles of the catalysts produced under
ð7Þ different final calcination temperatures (850–1000 °C) and dura-
tions (2 and 5 h). For all temperatures and durations, only a max-
imum of 1.3% weight loss was detected. This demonstrates that
the final calcination process has a significant effect on the thermal
stability of the catalyst. For the 2-h calcined catalysts, the thermal
ð8Þ stability increased (i.e. weight loss was minimised) as the calcina-
tion temperature was increased from 850 to 950 °C. An increase in
the calcination temperature enhances decomposition rate and
activity such as CaCO3 decomposition into active CaO which occurs
ð9Þ
between 825 and 900 °C. As reported previously, CaO displayed a
high thermal stability at any temperature due to its unique chem-
ical bonds [17]. Thus, increasing the calcination temperature above
850 °C could improve the catalyst thermal stability. However, an
increase in final calcination temperature to 1000 °C caused a slight
reduction in thermal stability. Calcination temperatures that are
too high could possibly damage the residual hydroxyl groups
ð10Þ bonded to the oxide lattice and cause defects in the catalyst struc-
ture [32]. For the 5-h calcined catalysts, an inconsistent observa-
tion is seen. The thermal stability did not exhibit an increasing
trend with temperature. The observed trend followed the order
of 950 > 1000 > 850 > 900 °C. This inconsistent trend could be due
to calcination duration being more dominant than temperature.
When calcination duration is prolonged, the thermal stability
ð11Þ reaches the maximum thermal stability regardless of the tempera-
ture used. Although an inconsistent trend is observed, it must be
borne in mind that the weight loss differences between each of
the cases were negligible (<0.2%).
 W.W.S. Ho et al. / Energy Conversion and Management xxx (2014) xxx–xxx 5

(a) (b)

Fig. 1. TGA profiles of calcined catalysts for (a) 2 h and (b) 5 h.

Comparing between the two different calcination durations, for
temperatures 850 and 950 °C, the TGA profiles obtained were al-
most identical. Only at 1000 °C, the increase in calcination duration
from 2 to 5 h promoted thermal stability as the weight loss was re-
duced by 1%. Generally, the increase in calcination duration did not
significantly improve the catalyst stability. However, due to the
longer calcination process, the energy consumption increased.
The improvement in catalyst stability observed was minimal, with
a maximum weight loss of 0.4%. Therefore, from the TGA profiles, it
could be concluded that calcination at 850 °C for 2 h was sufficient
to produce a high thermal stability catalyst.
3.1.3. Scanning electron microscopy (SEM) and energy dispersive X-
ray spectroscopy (EDS)
Fig. 2 shows the morphological characteristics of the calcined
catalysts under different temperatures and durations as captured
by the SEM micrographs. In general, all the catalysts exhibited sim-
ilar spongy and porous structures indicating highly adsorbent sur-
faces. All particles shown have uniform distribution of miniature
agglomerates with irregular shapes. These agglomerated structures
of the calcined catalysts indicated the formation of metal oxides
upon heat treatment [21]. As shown in Fig. 2(a)–(d), for the 2-h cal-
cined catalysts, the morphological sizes of the particles reduced as
the calcination temperature increased from 850 to 1000 °C. As a
result, the surface areas of the catalyst increased. For the 5-h cal-
cined catalysts as presented in Fig. 2(e)–(h), the morphological
sizes remained constant with the increase in calcination tempera-
ture. Again, as with the TGA profiles, the effect of calcination dura-
tion was more prominent than the effect of calcination
temperature. Calcination at high temperature for prolonged dura-
tion could possibly damage the agglomerated structure of the cat-
alyst. This can be seen from Fig. 2(h) which shows that the catalyst
calcined at 1000 °C for 5 h has bigger unevenly formed and distrib-
uted crystal chunks structure instead of a finer and more uniform
structure. This observation agrees well with Rashidi et al. [33] who
reported that too high a calcination temperature can damage the
catalyst structure. Once more, the increase in calcination duration
does not significantly improve the morphological characteristics of
the calcined catalyst, and therefore, calcination for 2 h was deemed
sufficient to produce a highly adsorbent catalyst. However, a finer
morphology size could be observed as the calcination temperature
increased to 950 °C. Overall, the SEM micrographs indicated that
950 °C was the most suitable calcination temperature to obtain
the finest porous catalyst.
Table 2 summarises the EDS elemental analysis for the devel-
oped catalysts. Since the data presented are considered semi-quan-
titative, all associated discussion using the elemental weight
percentages are relying on the figures to obtain valid trends when

(a) (b) (c) (d)

50 µm 50 µm 50 µm 50 µm

(e) (f) (g) (h)

50 µm 50 µm 50 µm 50 µm

Fig. 2. SEM images (2000 magnification) of catalysts calcined at (a) 850 °C, (b) 900 °C, (c) 950 °C and (d) 1000 °C for 2 h; (e) 850 °C, (f) 900 °C, (g) 950 °C and (h) 1000 °C for
5 h.

Please cite this article in press as: Ho WWS et al. Evaluation of palm oil mill fly ash supported calcium oxide as a heterogeneous base catalyst in biodiesel
synthesis from crude palm oil. Energy Convers Manage (2014), http://dx.doi.org/10.1016/j.enconman.2014.03.061
6 W.W.S. Ho et al. / Energy Conversion and Management xxx (2014) xxx–xxx

parameters are changed. The actual magnitudes must be treated

with a certain degree of conservatism. The analysis shows the
abundance of both Ca and oxygen (O) in CaCO3 before and after
Calcined
catalyst
final calcination. The Ca concentration increased upon calcination,
1000 °C

2.685
with more than 50 wt.% Ca being detected when CaCO3 was
5h
11
42

39
68
1
1
6
–
–
calcined at 800 °C. Furthermore, the elemental analysis for the
developed catalysts indicated that Ca was adsorbed by the ashes,
as high concentrations of Ca ranging from 34 to 38 wt.% were
Calcined
catalyst
1000 °C

noticeable in the catalysts. Several other metals, namely potassium

1.443
(K), magnesium (Mg), aluminium (Al), phosphorus (P) and silicon
2h
21
39

35
60
–
1
4
–
–
(Si) are the major components of the ashes. For the longer calcina-
tion duration of 5 h, a general increasing trend in Ca concentration
was observed as the temperature increased. Conversely, for the
Calcined
catalyst

shorter calcination duration of 2 h, a reverse trend was seen. In

950 °C

3.399
short, the final calcination conditions of 900 and 1000 °C for 5 h re-
5h
11
44

37
63
1
1
6
–
–

sulted in the highest Ca concentrations of 39 wt.% for both cases.

Nevertheless, the results also indicated that the catalyst calcined
at 850 °C for 2 h had comparatively high Ca concentration
Calcined
catalyst
950 °C

(38 wt.%) with respect to the catalysts developed under higher

5.026
2h

calcination temperatures and longer durations.

13
43

35
71
1
1
7
–
–

3.1.4. X-ray Diffraction (XRD)

The powder XRD patterns of the developed catalysts are de-
Calcined
catalyst
900 °C

5.753

picted in Fig. 3(a)–(h). Fig. 3(a) reveals only the CaO crystalline
5h

45

39
78

peaks indicating that the catalyst developed under this condition

8

1
1
6
–
–

had high CaO content. The characteristic peaks are observed at

2h = 32.32°, 37.48°, 53.95°, 64.31° and 67.47°. Peaks due to mullite
Calcined

(mainly aluminium oxide (Al2O3) and silicon oxide (SiO2)) and

catalyst
900 °C

4.754

dicalcium disilicate (Ca2SiO4) can be seen in Fig. 3(b)–(h). These

2h

47

36
75
7

1
1
8
–
–

combined peaks are due to the strong impregnation of active

CaO within the palm oil mill fly ash. The mullite peaks representing
the major ash components are observed at 2h = 26.7°, 39.5° and
Calcined
catalyst

50.7°. Meanwhile, the Ca2SiO4 crystalline phase is observed at

850 °C

6.399

2h = 30.9° for all catalysts except for the catalyst calcined at

5h
13
46

34
73
1
1
5
–
–

850 °C for 2 h and 900 °C for 5 h. The presence of Ca2SiO4 suggested

that Ca2SiO4 hydrate formed through the reaction between SiO2
and CaO in the presence of water. The Ca2SiO4 is an impurity and
Calcined
catalyst
850 °C

its formation should be avoided. Thus, the XRD results indicated

3.539
2h

that the catalysts calcined at 850 °C for 2 h and 900 °C for 5 h were
15
39

38
83
1
1
6
–
–
Elemental compositions, CaO concentrations and BET surface areas of the developed catalysts.

the best since they did not contain Ca2SiO4. Table 2 also lists the
CaO concentrations of the developed catalysts. The CaO concentra-
Calcined CaCO3 Uncalcined catalyst
after metal loading

tion of CaCO3 increased from 70% to 88 wt.% upon calcination at

800 °C. The uncalcined catalyst had a CaO concentration of only
56 wt.%. However, upon calcination under different conditions,
the CaO concentration significantly improved up to a maximum
0.485

of 83 wt.%. Calcination at 850 °C for 2 h and 900 °C for 5 h were

44

18
56
30

1
1
6
–
–

the best two conditions for maximum CaO concentrations. The

CaO concentrations obtained from the XRD analysis were in good
agreement with the Ca concentrations obtained from the EDS
at 800 °C

analysis.
9.63
43

51
88
6

–
–
–
–
–

3.1.5. Brunauer–Emmett–Teller (BET) surface area analysis

Catalyst type
Uncalcined

Table 2 also lists the BET surface areas of the catalysts devel-
oped under different calcination conditions. The surface area of
CaCO3

3.567 0.123

CaCO3 increased significantly from 0.123 to 9.63 m2 g1 upon cal-

16

34
50

70
–
–
–
–
–

cination at 800 °C. Likewise, the surface area significantly in-

creased from 0.485 m2 g1 for the uncalcined catalyst to a
ash
Fly

11
48

18

10

maximum of 6.399 m2 g1 for the calcined catalyst at 850 °C for

5
1

4
–
Mg

5 h. The catalysts calcined under different calcination tempera-

Ca
Al
Si
O

K
C

tures and durations displayed a wide range of surface areas from

CaO concentration

1.443 to 6.399 m2 g1. This indicated that the calcination condi-

BET surface area
composition

tions influenced the formation of pores within the catalysts. For

(m2 g1)
Elemental

the 2-h calcined catalysts, the surface areas increased as the tem-
(wt.%)

(wt.%)

perature increased up till 1000 °C when an unexpected drop in sur-

Table 2

face area occurred. Meanwhile, for the 5-h calcined catalysts, the
surface areas reduced as the temperature increased. The measured

Please cite this article in press as: Ho WWS et al. Evaluation of palm oil mill fly ash supported calcium oxide as a heterogeneous base catalyst in biodiesel
synthesis from crude palm oil. Energy Convers Manage (2014), http://dx.doi.org/10.1016/j.enconman.2014.03.061
 W.W.S. Ho et al. / Energy Conversion and Management xxx (2014) xxx–xxx
(a)
Intensity (counts)
(b)
(c)
(d)
2θ
Fig. 3. Powder XRD patterns of catalysts calcined at (a) 850 °C, (b) 900 °C, (c) 950 °C and (d) 1000 °C for 2 h; (e) 850 °C, (f) 900 °C, (g) 950 °C and (h) 1000 °C for 5 h
[characteristic peaks due to CaO (H), mullite (j) and dicalcium silicate (N)].
surface area trends are on the whole consistent with the SEM
observations of reduced morphological sizes with increased calci-
nation temperatures for the 2-h cases. For the 5-h cases, as temper-
ature increased, the surface areas reduced, but the morphological
sizes of the particles remained constant. The evidence seen here
points to the longer calcination duration damaging the pores at
the catalyst surfaces. This resulted in the blockage of the pores
and subsequently, the reduction in surface areas although the mor-
phological sizes remained the same. The catalysts with the lowest
surface areas were obtained from calcinations at 1000 °C for both 2
and 5 h. This could also possibly be attributed to the blocking of
pores within the ash supported CaO catalyst. An inversely propor-
tional trend between surface area and Ca concentration is also ob-
served for both the 2-h and 5-h calcined catalysts. It has been
reported that when CaO concentration increases, the mesopore
volume of the catalyst decreases which leads to lower surface area
[34].
3.2. Transesterification of CPO
3.2.1. Optimisation of catalyst final calcination conditions
Table 3 lists the biodiesel yields from the transesterification
reactions catalysed by waste fly ash supported CaO catalysts devel-
oped under different final calcination conditions. The catalysts cal-
cined for 2 h resulted in higher biodiesel yields at lower calcination
temperatures (850 and 900 °C). Conversely, the catalysts calcined
for 5 h had higher yields at higher calcination temperatures (950
and 1000 °C). At lower temperatures, the increase in calcination
duration from 2 to 5 h resulted in the reduction of the average
yields (of all catalyst loadings) from 58.18% to 51.77%, and
58.09% to 54.09% for 850 and 900 °C, respectively. In contrast, at
higher temperatures, the increase in calcination duration improved
the average yields from 56.93% to 62.89%, and 54.67% to 57.35% for
Please cite this article in press as: Ho WWS et al. Evaluation of palm oil mill fly ash supported calcium oxide as a heterogeneous base catalyst in biodiesel
synthesis from crude palm oil. Energy Convers Manage (2014), http://dx.doi.org/10.1016/j.enconman.2014.03.061
7
(e)
(f)
(g)
(h)
950 and 1000 °C, respectively. The increase in calcination duration
was considered undesirable as it did not alter the biodiesel yield
significantly, but would increase the energy consumption. For the
2-h calcined catalysts, as the temperature increased from 850 to
1000 °C, the average yields decreased from 58.18% to 54.67%.
Meanwhile, for the 5-h calcined catalysts, an increase in average
yields from 51.77% to 62.89% was obtained as temperature in-
creased from 850 to 950 °C. This was followed by a decrease to
57.35% at 1000 °C. The results suggested that the calcination
temperature exerted a greater influence on the average yield than
the calcination duration. This is in line with Ngamcharussrivichai
et al.’s [32] work in which it was reported that the increase in cal-
cination temperature exerted a much greater influence on the CO2
dissociation rate as compared to the calcination duration. Since the
dissociation of CO2 progresses from the outer surface into the par-
ticle, a CO2 film is formed at the surface leading to the recarbona-
tion of CaO to CaCO3. This recarbonation negatively affects the
catalytic performance of the catalyst.
From Table 3, it can be seen that the catalyst calcined at 850 °C
for 2 h resulted in the highest biodiesel yield of 75.73%. This is con-
sistent with the characterisation results which indicated that the
catalyst calcined at 850 °C for 2 h displayed high CaO and Ca con-
centrations. Thus, the catalyst had high basicity despite having a
lower surface area. Additionally, this catalyst displayed a high ther-
mal stability as inferred from its TGA. Hence, it could be concluded
that a calcination temperature of 850 °C was sufficient to produce
an effective catalyst. This temperature also matched the decompo-
sition temperature of CaCO3. Upon calcination to 850 °C, CaO is
formed, which has many defects in its crystal structure. These de-
fects favour the formation of calcium methoxide, which is a surface
intermediate in transesterification. Generally, for 2-h calcined cat-
alyst, when the calcination temperature exceeded 900 °C, the cat-
alytic activity gradually decreased despite the increase in BET
8 W.W.S. Ho et al. / Energy Conversion and Management xxx (2014) xxx–xxx

Table 3 sites available for methanol to form methoxide anion. However,

Biodiesel yields from transesterification reactions catalysed by developed catalysts. both yield and conversion decreased when the catalyst loading ex-
Calcination Calcination Catalyst loading Biodiesel ceeded 6 wt.%, down to 49.41% and 84.82%, respectively at 8 wt.%
temperature (°C) duration (h) (wt.%) yield (%) loading. This could possibly be due to the emulsion formed by
850 2 2 51.22 the high catalyst content leading to soap formation and reduced
4 55.18 biodiesel production. Furthermore, at high catalyst loading, the ef-
6 75.73 fect of mass transfer intensification becomes more dominant than
8 50.59
5 2 43.96
the effect of increased catalyst amount [36]. The high reaction mix-
4 47.64 ture viscosity would result in poor mass diffusion within the meth-
6 63.53 anol-oil-heterogeneous catalyst system [37].
8 51.94 Fig. 4(b) depicts the effects of methanol to oil molar ratio on
900 2 2 52.81 biodiesel yield and FAME conversion. Since transesterification is
4 58.38 a reversible reaction, sufficient methanol must be present during
6 70.59
the transesterification process to trigger and drive the reaction for-
8 50.59
5 2 45.77 ward. By increasing the methanol to oil molar ratio from 6:1 to 9:1,
4 46.82 the biodiesel yield and FAME conversion increased from 73.06% to
6 65.41 73.42%, and 82.41% to 94.64%, respectively. Further increment to
8 58.35 12:1 improved both the yield and conversion, whereby a maxi-
950 2 2 50.10 mum of 75.73% yield and 98.30% conversion were achieved. How-
4 55.27 ever, a reduction in yield from 75.73% to 54.12% was observed
6 63.53
8 58.82
when the molar ratio exceeded 12:1. Meanwhile, the conversion
5 2 51.34 dropped to almost constant at 97%. A methanol to oil molar ratio
4 71.76 of 12:1 was deemed to be the best ratio offering the highest yield.
6 73.18 Above this ratio, the higher methanol concentration leads to de-
8 55.29
creased yields due to the increased methanol–glycerol solubility
1000 2 2 49.62 which interferes with the glycerol separation [37]. The presence
4 71.29
of the polar hydroxyl group in methanol promotes emulsification
6 51.17
8 46.59 of the product [38].
5 2 54.41 The effects of reaction time on biodiesel yield and FAME conver-
4 59.76 sion can be seen in Fig. 4(c). At 1 and 2 h, experimental tests
6 61.89 showed that the conversions were very low at below 50% (data
8 53.34
not shown). This was caused by the slow reaction rate due to dis-
Transesterification conditions: Methanol to oil molar ratio 12:1; temperature 60 °C, persion and mixing between the methanol and oil. Generally,
time 3 h; stirring speed 700 rpm. increasing the reaction time promotes the transesterification pro-
cess towards completion. A maximum of FAME conversion of
98.30% was observed at the end of 3 h and thereafter it remained
surface area. The effects on the catalytic activity were not merely a approximately constant at 97.50%. The slight drop in conversion
result of the change in the catalyst surface area, but also of the could be partly associated to the formation of glycerol under longer
transformation within the crystal structure. The structure became duration [39]. In contrast, the biodiesel yield decreased when the
more intact with the rise in calcination temperature. The coordina- reaction time exceeded 3 h. This was because a longer duration en-
tion of Ca2+ and O2 on the catalyst surface changed which varied hanced the hydrolysis of esters (reversed transesterification)
the amount and strength of basic sites thereby significantly affect- resulting in the loss of esters as well as causing more fatty acids
ing the catalytic performance [13]. The findings obtained in this to form soap [40]. The results here indicated that the suitable time
work indicated that the overall catalytic activity does not depend required for achieving the highest amount of biodiesel yield was
solely upon its surface area or other factors such as basicity, 3 h.
porosity and morphology size, but the combined effects of these Fig. 4(d) illustrates the effects of reaction temperature on bio-
multiple attributes [35]. diesel yield and FAME conversion. In a methanol-oil-heteroge-
neous catalyst system, the reaction rate decreases because of
3.2.2. Effects of transesterification reaction parameters diffusional resistance between the liquid–liquid–solid phases.
To investigate the effects of transesterification reaction param- However, the reaction rate can be accelerated with higher reaction
eters on biodiesel yield and FAME conversion, the catalyst devel- temperatures. From Fig. 4(d), it could be observed that a low reac-
oped using a calcination temperature of 850 °C and duration of tion temperature of 45 °C was sufficient to achieve a high FAME
2 h was employed in a single factor parametric study. Fig. 4(a) conversion of 97.09%. As the reaction temperature increased from
shows the effects of catalyst loading on biodiesel yield and FAME 45 to 65 °C, the FAME conversion improved slightly from 97.09%
conversion. Both biodiesel yield and FAME conversion increased to 98.59% as higher temperatures increase reaction rates. However,
as the catalyst loading increased. The biodiesel yield increased the highest FAME conversion might not be necessarily accompa-
from 51.22% to 75.73% while the FAME conversion improved nied by the highest yield. A decreasing trend of yield was observed
slightly from 96.57% to 98.30% when the catalyst loading increased with increasing reaction temperature. At the highest temperature
from 2 to 6 wt.%. Thus, it could be seen that the developed waste of 65 °C, the highest FAME conversion of 98.59% was attained to-
fly ash supported CaO catalyst positively affected the yield and gether with the lowest yield of 73.53%. At this temperature, the
conversion due to its basicity. At low catalyst loading of 2 wt.%, boiling reaction mixture formed a large number of bubbles, which
although the FAME conversion exceeded the minimum ester con- inhibited the reaction especially on the three-phase interfaces [36].
tent requirement, the low biodiesel yield of 51.22% was undesir- Furthermore, higher temperatures can accelerate the side saponifi-
able. The maximum biodiesel yield reached 75.73% with a high cation of triglycerides [40]. The combination of these two effects
FAME conversion of 98.30% when 6 wt.% of catalyst loading was likely led to the decrease in the final biodiesel yield. Nevertheless,
used. This increment could be attributed to the increase in basic the differences in both yield and conversion are small at 1.5% and

Please cite this article in press as: Ho WWS et al. Evaluation of palm oil mill fly ash supported calcium oxide as a heterogeneous base catalyst in biodiesel
synthesis from crude palm oil. Energy Convers Manage (2014), http://dx.doi.org/10.1016/j.enconman.2014.03.061
 W.W.S. Ho et al. / Energy Conversion and Management xxx (2014) xxx–xxx 9

(a) (b)

(c) (d)

Fig. 4. Effects of (a) catalyst loading, (b) reaction temperature, (c) methanol to oil molar ratio and (d) reaction time on biodiesel yield and FAME conversion under nominal
parametric settings.
FAME conversion (%)

-ln (1-X)

(a) Reaction time (min) (b) Reaction time (min)

(c)
Fig. 5. (a) FAME conversion versus time at 6 wt.% catalyst loading, 12:1 methanol to oil molar ratio and 700 rpm stirring speed under different reaction temperatures, (b)
ln(1  X) versus t at different reaction temperatures and (c) Arrhenius plot of ln k versus 1/T  103.

Please cite this article in press as: Ho WWS et al. Evaluation of palm oil mill fly ash supported calcium oxide as a heterogeneous base catalyst in biodiesel
synthesis from crude palm oil. Energy Convers Manage (2014), http://dx.doi.org/10.1016/j.enconman.2014.03.061
10 W.W.S. Ho et al. / Energy Conversion and Management xxx (2014) xxx–xxx

6.13%, respectively. This indicated that temperature was not a Table 4

dominant factor when the reaction time was long. A reaction with Rate constants k and correlation coefficients R2 for the transesterification of CPO at
different temperatures.
higher temperature can attain the maximum FAME conversion or
equilibrium state in a shorter time, but will remain constant there- Temperature (°C) Rate constant, k (min1) R2
after. With this in mind, a temperature of 45 °C was sufficient to 45 0.0041 0.944
achieve a yield of 79.76% and an FAME conversion of 97.09%. Addi- 50 0.0063 0.916
tional increase in reaction temperature above this was not neces- 55 0.007 0.936
60 0.0082 0.956
sary in the case of the current work considering the increase in 65 0.0118 0.942
energy consumption.

3.3. Kinetic modelling 3.4. Biodiesel physicochemical properties

Fig. 5(a) illustrates that FAME conversion increases as reaction
time increases. FAME conversion was slow in the initial stage. This
was followed by a more rapid increase and finally, the conversion
reached equilibrium at 180 min. An increasing trend in the FAME
conversion with temperature was observed since higher tempera-
tures favoured the reaction. However, the highest FAME conver-
sion obtained at 65 °C did not necessarily produce the highest
biodiesel yield. On the contrary, the lowest yield was obtained
due to saponification which occurred at high temperatures [41].
By plotting ln(1  XME) versus t as shown in Fig. 5(b), strong
linear correlations were obtained which supported the hypothesis
of pseudo-first order reaction. Table 4 summarises the k values for
the temperatures studied and the corresponding correlation coeffi-
cients (R2) obtained from the linear fitting. Fig. 5(c) shows that
string linearity (R2 = 0.95) could be observed between ln k and 1/
T across the range of temperatures studied. The Ea determined
from the slope of the line was 42.56 kJ mol1, while the ko calcu-
lated from the intercept was 42192.6 min1. As such, the Arrhenius
equation for this heterogeneous catalysed reaction could be writ-
ten as:
The key biodiesel physicochemical properties are tabulated in
Table 5. The measured ester and linolenic acid contents, kinematic
viscosity, density, flash point, acid value, HHV and CN are within
the limits specified by the European Standard EN 14214 [24]. It
is important that the properties meet the requirements set since
they impact both the structure and performance of diesel engines.
For instance, to prevent coking on the injectors, kinematic viscosity
is a critical consideration. The measured kinematic viscosity at
40 °C was slightly below the upper limit of 6.0 mm2 s1. The mea-
sured density at 15 °C was slightly above the lower limit of
860 kg m3. The flash point of the biodiesel was 144 °C. The high
flash point value ensures that the fuel is safe in terms of storage
and transportation. The acid value was determined to be
0.42 mg KOH g1 which was below the upper limit of
0.5 mg KOH g1. An upper limit is imposed on acid value to avoid
corrosion related problems. The HHV was measured as
38.1 MJ kg1 while the cetane number was predicted as 48.1. Both
of these properties are indicators of the quality of the fuel.
3.5. Catalyst reusability

ln k ¼ 10:65  5:1181=T ð12Þ 3.5.1. Optimisation of catalyst regeneration method

Catalyst recycling is an important step as it reduces the overall
where k is the rate constant (min1) and T is the temperature (K). biodiesel production cost. Prior to recycling, the spent catalyst

Table 5
Physicochemical properties of produced biodiesel.

Physicochemical property Units Test method Limits Produced biodiesel

Kinematic viscosity mm2 s1; 40 °C ASTM D 445 1.9–6.0 5.9
Density kg m3; 15 °C ASTM D 1298 860–900 865.3
Flash point °C ASTM D 93 130 minimum 144
Acid value mg KOH g1 ASTM D 664 0.5 maximum 0.42
Higher heating value MJ kg1 ASTM D 4868 35 minimum 38.1
Cetane number – ASTM D613 47 minimum 48.1
Ester content % (mol mol1) EN 14103 96.5 minimum 97.1
Linolenic acid content % (mol mol1) EN 14103 12 maximum 0.18

Table 6
CaO concentrations and elemental compositions of fresh, spent and regenerated catalysts.

Fresh catalyst Spent catalyst Regenerated catalyst

Method 1a Method 2b Method 3c
CaO concentration (wt.%) 83 55 67 70 74
Elemental composition (wt.%) C 15 39 30 14 16
O 39 40 36 36 36
Mg 1 1 1 1 1
Al 1 – 2 1 1
Si 6 3 3 13 10
K – – – 3 2
Ca 38 17 28 32 34
a
Methanol washing.
b
Recalcination without methanol washing.
c
Methanol washing followed by recalcination.

Please cite this article in press as: Ho WWS et al. Evaluation of palm oil mill fly ash supported calcium oxide as a heterogeneous base catalyst in biodiesel
synthesis from crude palm oil. Energy Convers Manage (2014), http://dx.doi.org/10.1016/j.enconman.2014.03.061
 W.W.S. Ho et al. / Energy Conversion and Management xxx (2014) xxx–xxx 11

needs to be regenerated as failure to do so would deteriorate the

catalytic performance [14]. Table 6 lists the XRD and EDS results
of the fresh, spent and regenerated catalysts for the three different
regeneration methods. The CaO concentration decreased from
83 wt.% to 55 wt.% while the Ca concentration decreased from
38 wt.% to 17 wt.% after the first usage. From Table 6, methanol
washing followed by recalcination 850 °C for 2 h was identified
as the best regeneration method in which the CaO and Ca concen-
trations were regenerated to 74 wt.% and 34 wt.%, respectively.
These concentration were only 9 wt.% and 4 wt.% less than the ori-
ginal CaO and Ca concentrations, respectively. Recalcining the cat-
alyst converted the CaCO3 back to active CaO by removing CO2 and
other impurities. The element Al concentration after recalcining
was also acceptable. The Si concentration (major element of ash)
meanwhile increased upon recalcination. Fig. 6. Biodiesel yields and FAME conversions in reusability study (6 wt.% catalyst
loading, 12:1 methanol to oil molar ratio, 45 °C temperature, 3 h time and 700 rpm
stirring speed).
3.5.2. Reusability cycles
Catalyst reusability was tested under the optimum transesteri-
fication conditions of 6 wt.% catalyst loading, 12:1 methanol to oil
molar ratio, 45 °C reaction temperature, 3 h reaction time and Table 7
700 rpm stirring speed. After each cycle, the spent catalyst was CaO and Ca concentrations and BET surface areas of fresh and recycled catalysts.
regenerated using methanol washing followed by recalcination at Catalyst CaO concentration Ca concentration BET surface
850 °C for 2 h. As shown in Fig. 6, the yield and conversion dropped (wt.%) (wt.%) area (m2 g1)
to 38.93% and 53.79%, respectively in the fourth cycle. Biodiesel Fresh 83 38 3.539
was not produced in the cycle after that. Both the yield and conver- 2nd cycle 74 34 3.530
sion obtained in the fourth cycle were unacceptably low. The 3rd cycle 66 28 3.489
regenerated catalyst was active for three consecutive cycles 4th cycle 60 15 3.435
5th cycle 56 10 3.391
whereby the minimum yield and conversion were 60.75% and
70.72%, respectively. The decrease in yield could be attributed to
catalyst deterioration due to poisoning by glycerol and soap pres-
ent in the reaction mixture [36]. Dissolution of Ca from CaO to the transesterification cycle, the CaO concentration decreased from
alcoholic phase also caused the deactivation of active sites [42]. 83 wt.% to 74 wt.% which suggested that dissolution of CaO parti-
The presence of calcium glyceroxide in the spent catalyst could cles from the catalyst surface occurred [34]. After the third cycle,
possibly be a factor in its reduced reusability as reported by Kouzu the CaO concentration drastically fell to 60 wt.%, which was consis-
et al. [43]. Calcium glyceroxide is formed through the combination tent with the low observed catalytic activity. The decreases in yield
of CaO and glycerol [43]. Due to the low catalytic performance in and conversion after each cycle were also supported by the
the subsequent cycles, the yield could possibly be increased with decreasing trend shown by the Ca concentration from 38 to
the use of higher catalyst loading and longer reaction time. 15 wt.%.
Although ideally a catalyst should be reused as many times as pos- The fresh fly ash based-catalyst had a surface area of
sible, the limited number of reusability cycles with this fly ash sup- 3.539 m2 g1. The recycled catalysts had approximately the same
ported CaO is compensated by the fact that palm oil mill fly ash is a surface areas as compared to the fresh one. From this, it could be
waste product. inferred that the catalytic performance of the recycled catalysts
was linked more to the basicity property than the surface area.
3.5.3. Characterisation of the recycled catalysts Fig. 7 illustrates the SEM micrographs of the catalysts captured be-
The CaO and Ca concentrations of the recycled catalysts were fore and after the transesterification cycle. The fresh catalyst had
determined using XRD and EDS, respectively. The BET surface area uniform distribution of miniature agglomerates with irregular
was also measured. Table 7 shows the CaO and Ca concentrations shapes. After the reaction, the spent catalyst appeared to be made
and the BET surface areas of the fresh and recycled catalysts. The up of more condensed mass. Mootabadi et al. [42] have suggested
fresh fly ash based-catalyst had a high CaO concentration of that physical changes occurred after reaction and new active sur-
83 wt.% and a Ca concentration of 38 wt.%. The CaO concentration face morphological sites were formed, which had significant effect
decreased considerably after each subsequent cycle. After the first on catalytic activity.

(a) (b) (c) (d)

50 µm 50 µm 50 µm 50 µm

Fig. 7. SEM images (2000 magnification) of catalysts from (a) 1st cycle, (b) 2nd cycle, (c) 3rd cycle and (d) 4th cycle.

Please cite this article in press as: Ho WWS et al. Evaluation of palm oil mill fly ash supported calcium oxide as a heterogeneous base catalyst in biodiesel
synthesis from crude palm oil. Energy Convers Manage (2014), http://dx.doi.org/10.1016/j.enconman.2014.03.061
12 W.W.S. Ho et al. / Energy Conversion and Management xxx (2014) xxx–xxx
4. Conclusions
This study focussed on the improved development and evalua-
tion of a palm oil mill fly ash supported CaO catalyst. This hetero-
geneous base catalyst was used in the transesterification of CPO
with methanol to yield biodiesel. The optimum final calcination
conditions with 45 wt.% loading of calcined CaCO3 were 850 °C
and 2 h. Under these conditions, 75.73% of biodiesel yield was ob-
tained. It was determined that transesterification conditions of
6 wt.% catalyst loading, 12:1 methanol to oil molar ratio, 45 °C
reaction temperature, 3 h reaction time and 700 rpm stirring speed
resulted in biodiesel yield and FAME conversion of 79.76% and
97.09%, respectively. The kinetic study suggested that the hetero-
geneous catalysed transesterification was a pseudo-first order
reaction. The Ea of the reaction was determined as 42.56 kJ mol1.
Key biodiesel physicochemical properties including ester and lino-
lenic acid contents, kinematic viscosity, density, flash point, acid
value, HHV and cetane number were within the limits specified
in EN 14214. The developed catalyst could also be reused feasibly
up to three consecutive cycles after regeneration using methanol
washing followed by recalcination at 850 °C for 2 h. Overall, the
potential of this low-cost heterogeneous base catalyst has been
demonstrated for transesterification applications.
Acknowledgements
This work was supported by the Ministry of High Education
(MOHE), Malaysia as well as the Faculty of Engineering at The Uni-
versity of Nottingham Malaysia Campus. The authors would also
like to express gratitude to Havys Oil Mill Sdn. Bhd. for providing
the crude palm oil and Seri Ulu Langat Palm Oil Mill for the supply
of palm oil mill boiler ash.
References
[1] Sharma YC, Singh B, Korstad J. Latest developments on application of
heterogeneous basic catalysts for an efficient and eco friendly synthesis of
biodiesel: a review. Fuel 2011;90:1309–24.
[2] Granados ML, Alonso DM, Alba-Rubio AC, Mariscal R, Ojeda M, Brettes P.
Transesterification of triglycerides by CaO: increase of the reaction rate by
biodiesel addition. Energy Fuels 2009;23:2259–63.
[3] Helwani Z, Othman MR, Aziz N, Fernando WJN, Kim J. Technologies for
production of biodiesel focusing on green catalytic techniques: a review. Fuel
Process Technol 2009;90:1502–14.
[4] Boey P-L, Gaanty PM, Shafida AH. Performance of calcium oxide as a
heterogeneous catalyst in biodiesel production: a review. Chem Eng J
2011;168:15–22.
[5] Kouzu M, Hidaka J-S. Transesterification of vegetable oil into biodiesel
catalysed by CaO: a review. Fuel 2012;93:1–12.
[6] Dias JM, Alvim-Ferraz MCM, Almeida MF, Diaz JDM, Polo MS, Utrilla JR.
Biodiesel production using calcium manganese oxide as catalyst and different
raw materials. Energy Convers Manage 2013;65:647–53.
[7] Kouzu M, Kasuno T, Tajika M, Sugimoto Y, Yamanaka S, Hidaka J. Calcium oxide
as a solid base catalyst for transesterification of soybean oil and its application
to biodiesel production. Fuel 2008;87:2798–806.
[8] Singh Chouhan AP, Sarma AK. Modern heterogeneous catalysts for biodiesel
production: a comprehensive review. Renew Sust Energy Rev
2011;12:4378–99.
[9] Watkins RS, Adam FL, Wilson K. Li–CaO catalysed tri-glyceride
transesterification for biodiesel applications. Green Chem 2004;4:335–40.
[10] MacLeod CS, Harvey AP, Lee AF, Wilson K. Evaluation of the activity and
stability of alkali-doped metal oxide catalysts for application to an intensified
method of biodiesel production. Chem Eng J 2008;135:63–70.
[11] Alonso DM, Mariscal R, Granados ML, Maireles-Torres P. Biodiesel preparation
using Li/CaO catalysts: activation process and homogeneous contribution.
Catal Today 2009;143:167–71.
[12] Ngamcharussrivichai C, Totarat P, Bunyakiat K. Ca and Zn mixed oxide as a
heterogeneous base catalyst for transesterification of palm kernel oil. Appl
Catal A 2008;341:77–85.
Please cite this article in press as: Ho WWS et al. Evaluation of palm oil mill fly ash supported calcium oxide as a heterogeneous base catalyst in biodiesel
synthesis from crude palm oil. Energy Convers Manage (2014), http://dx.doi.org/10.1016/j.enconman.2014.03.061
[13] Zhu H, Wu Z, Chen Y, Zhang P, Duan S, Liu X, et al. Preparation of biodiesel
catalysed by solid super base of calcium oxide and its refining process. Chin J
Catal 2006;27:391–6.
[14] Boey P-L, Gaanty PM, Shafida AH. Biodiesel production via transesterification
of palm olein using waste mud crab (Scylla serrata) shell as a heterogeneous
catalyst. Bioresour Technol 2009;100:6362–8.
[15] Boey P-L, Maniam GP, Hamid SA, Ali DMH. Utilisation of waste cockle shell
(Anadara granosa) in biodiesel production from palm olein: optimisation using
response surface methodology. Fuel 2011;90:2353–8.
[16] Oliveira DA, Benelli P, Amante ER. A literature review on adding value to solid
residues: egg shells. J Clean Prod 2013;46:42–7.
[17] Ho WWS, Ng HK, Gan S. Development and characterisation of novel
heterogeneous palm oil mill boiler ash-based catalysts for biodiesel
production. Bioresour Technol 2012;1255:158–64.
[18] Sim JH, Kamaruddin AH, Bhatia S. The feasibility study of crude palm oil
transesterification at 30 °C operation. Bioresour Technol 2010;10:8948–54.
[19] Melero JA, Bautista LF, Morales G, Iglesias J, Sanchez-Vazquez R. Biodiesel
production from crude palm oil using sulfonic acid-modified mesostructured
catalysts. Chem Eng J 2010;161:323–31.
[20] Chakraborty R, Bepari S, Banerjee A. Transesterification of soybean oil
catalysed by fly ash and egg shell derived solid catalysts. Chem Eng J
2010;165:798–805.
[21] Hayyan A, Alam MdZ, Mirghani MES, Kabbashi NA, Nazashida NI, Siran YM,
et al. Sludge palm oil as a renewable raw material for biodiesel production by
two-step processes. Bioresour Technol 2010;101:7804–11.
[22] Atapour M, Kariminia H-R. Characterisation and transesterification of Iranian
bitter almond oil for biodiesel production. Appl Energy 2011;88:2377–81.
[23] Leung DYC, Guo Y. Transesterification of neat and used frying oil: optimisation
for biodiesel production. Fuel Process Technol 2006;87:883–90.
[24] EN 14214. Automotive fuels – Fatty Acid Methyl Esters (FAME) for diesel
engines – requirements and test methods; 2003.
[25] Lempers HEB, Sheldon RA. The stability of chromium in CrAPO-5, CrAPO-11,
and CrS-1 during liquid phase oxidations. J Catal 1998;175:62–9.
[26] Chipera SJ, Bish DL. Fitting full X-ray diffraction patterns for quantitative
analysis: a method for readily quantifying crystalline and disordered phases.
Adv Mater Phys Chem 2013;3:47–53.
[27] Issariyakul T, Dalai AK. Comparative kinetics of transesterification for biodiesel
production from palm oil and mustard oil. Can J Chem Eng 2012;90:342–50.
[28] EN 14103. Fat and oil derivatives – Fatty Acid Methyl Esters (FAME) –
determination of ester and linolenic acid methyl esters contents; 2003.
[29] Cheenkachorn K. Prediction properties of biodiesels using statistical models
and artificial neural network. Sust Energy Environ 2014;3003:176–9.
[30] Zhang L, Sheng B, Xin Z, Liu Q, Sun S. Kinetics of transesterification of palm oil
and dimethyl carbonate for biodiesel production at the catalysis of
heterogeneous base catalyst. Bioresour Technol 2010;101:8144–50.
[31] Birla A, Singh B, Upadhyay SN, Sharma YC. Kinetics studies of synthesis of
biodiesel from waste frying oil using a heterogeneous catalyst derived from
snail shell. Bioresour Technol 2012;106:95–100.
[32] Ngamcharussrivichai C, Nunthasanti P, Tanachai S, Bunyakiat. Biodiesel
production through transesterification over natural calciums. Fuel Process
Technol 2010;91:1409–15.
[33] Rashidi NA, Mohamed M, Yusup S. A study of calcination and carbonation of
cockle shell. World Acad Sci Eng Technol 2011;60:818–23.
[34] Witoon T, Bumrungsalee S, Vathavanichkul P, Palisakun S, Saisriyoot M,
Faungnawakij K. Biodiesel production from transesterification of palm oil with
methanol over CaO supported on bimodal meso-macroporous silica catalyst.
Bioresour Technol 2014;156:329–34.
[35] Li E, Rudolph V. Transesterification of vegetable oil to biodiesel over MgO-
functionalised mesoporous catalysts. Energy Fuels 2008;22:145–9.
[36] Liu X, Piao X, Wang Y, Zhu S. Calcium ethoxide as a solid base catalyst for the
transesterification of soybean oil to biodiesel. Energy Fuels 2008;22:1313–7.
[37] Kotwal MS, Niphadkar PS, Deshpande SS, Bokade VV, Joshi PN.
Transesterification of sunflower oil catalysed by flyash-based solid catalysts.
Fuel 2009;88:1773–8.
[38] Viriya-empikul N, Krasae P, Nualpaeng W, Yoosuk B, Faungnawakil K. Biodiesel
production over Ca-based solid catalyst derived from industrial wastes. Fuel
2012;92:239–44.
[39] Calcens J-M, Pouilloux Y, Barrault J. Selective etherification of glycerol to
polyglycerols over impregnated basic MCM-41 type mesoporous catalysts.
Appl Catal A 2002;227:181–90.
[40] Eevera T, Rajendran K, Saradha S. Biodiesel production process optimisation
and characterisation to assess the suitability of the product for varied
environmental conditions. Renew Energy 2009;34:762–5.
[41] Leung DYC, Guo Y. Transesterification of neat and used frying oil: optimization
for biodiesel production. Fuel Process Technol 2006;87:883–90.
[42] Mootabadi H, Salamatinia B, Bhatia S, Abdullah AZ. Ultrasonic-assisted
biodiesel production process from palm oil using alkaline earth metal oxides
as the heterogeneous catalysts. Fuel 2010;89:1818–25.
[43] Kouzu M, Kasuno T, Tajika M, Yamanaka S, Hidaka J. Active phase of calcium
oxide used as solid base catalyst for transesterification of soybean oil with
refluxing methanol. Appl Catal A 2008;334:357–65.