Fuel 98 (2012) 194–202

Contents lists available at SciVerse ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Biodiesel production based on heterogeneous process catalyzed

by solid waste coral fragment
Wuttichai Roschat a, Mattana Kacha a, Boonyawan Yoosuk b, Taweesak Sudyoadsuk a, Vinich Promarak a,⇑
a
Center for Organic Electronic and Alternative Energy (COEA), Department of Chemistry and Center for Innovation in Chemistry, Faculty of Science, Ubon Ratchathani University,
Warinchumrap, Ubon Ratchathani 34190, Thailand
b
National Metal and Materials Technology Center (MTEC), 114 Thailand Science Park, Phahonyothin Road, Klong 1, Klong Luang, Pathumthani 12120, Thailand

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

Article history: We present an environmentally friendly, efficient, simple heterogeneous transesterification process com-
Received 7 November 2011 bining palm oil, soybean oil, rice bran oil, and waste cooking oil with methanol to make a biodiesel and
Received in revised form 8 April 2012 glycerol by-product using calcined waste coral fragments in solid form as a catalyst. Under optimum
Accepted 9 April 2012
reaction conditions, the coral fragments calcined at 700 °C for 1 h; catalyst/oil ratio of 100 wt.%; metha-
Available online 22 April 2012
nol/oil molar ratio of 15:1; reaction temperature of 65 °C with a constant stirring are able to transesterify
oils to a biodiesel product with a FAME yield over 98% in 2 h. The catalyst can be easily separated from the
Keywords:
reaction mixture by pouring off the reaction solution and can be reused several times with consistent
Biodiesel
Glycerol
results. Biodiesel and glycerol products without any treatment and cleansing show a high quality product
Heterogeneous process in which fuel properties of the biodiesel meet all EN 14214 standards for bio-auto fuels.
Solid catalyst Ó 2012 Elsevier Ltd. All rights reserved.
Coral fragment

1. Introduction removal of these catalysts is costly, time consuming and generates

large amounts of waste water. In addition, crude glycerol by-prod-
Since the potential shortages and rise of petroleum prices, and ucts have very low value due to high impurities such as other cat-
ever-increasing environmental concerns, a search for alternative alysts, soap, methanol and water. Purifying the crude glycerol by
renewable fuels has drawn numerous intentions [1]. Biodiesel, a means of filtration, chemical additions, fractional vacuum distilla-
clean renewable fuel derived from biological sources, is a promis- tion, bleaching, deodoring, and ion exchange yields various com-
ing candidate for a diesel fuel substitute because it can be used mercial and technical grades, however, it is costly, especially for
in any diesel-engine vehicle without the need for modification medium and small biodiesel producers.
and produces fewer amounts of polluting emissions. Biodiesel is Therefore, there is a need for a milder, environmentally friendly
chemically a fatty acid methyl ester (FAME) and can be produced and integrated process producing cost-efficient biodiesel and high-
from a direct transesterification of vegetable oils or animal fat, quality glycerol. The use of a heterogeneous catalyst [5] is a key
where the corresponding triglycerides react with methyl alcohol technology to overcome such problems. The low end of pipe costs,
in the presence of a catalyst forming glycerol as a by-product energy savings, high efficiency, and selectivity are especially strong
[2,3]. Nowadays, the costs of the triglyceride starting material driving forces behind the development of new green and sustain-
and production process represent the majority of the price of bio- able technologies. Heterogeneous catalysts offer significant advan-
diesel. The high triglyceride quality is always in demand with food tages of simple isolation from the reaction, requiring no cleansing
producers. Therefore, the utilization of the glycerol by-product of with water, very little environmental pollution, and the possibility
biodiesel production would significantly improve the economic as- to be reused and recycled. Many of them have been investigated
pect of overall biodiesel production. Glycerol is one of the most including metal oxides [6], calcined hydrotalcites [7], zeolites [8],
versatile and valuable known chemicals and has a wide variety and anion exchanged resins [9]. However, good yields of biodiesel
of uses and applications such as food, pharmaceutical, and cos- in mild condition are achieved mainly by the catalysts in powder
metic, coating and other industries. Currently, homogeneous form [10]. Moreover, the scaling up of this process to an industrial
process catalyzed by sodium or potassium hydroxide is a common system with fixed based reactors would be impossible due to the
and efficient method for biodiesel production [4] though the high pressure drop occurring along the reactor, and removal of a
powder catalyst from biodiesel is also extremely difficult with
⇑ Corresponding author. Tel.: +66 4535 3400x4510; fax: +66 4535 3422x4510. batch reactor. Therefore, in order to develop a heterogeneous cat-
E-mail addresses: [email protected], [email protected] (V. Promarak). alyst with practical industrial applications, it is necessary to search

0016-2361/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.fuel.2012.04.009
 W. Roschat et al. / Fuel 98 (2012) 194–202 195

for a catalyst with high activity in particle, pellet or bulk solid form. was kept at room temperature for 24 h to attain a maximum water
To date, there is a rare report on high conversion of biodiesel pro- adsorption. The coral fragment was carefully removed from water
duction using bulk solids as catalyst [11,12]. and weighted. The amount adsorbed water was calculated from a
Among the metal oxides, calcium oxide (CaO) is one of the sol- weight difference, which was then converted to a total pore vol-
ids that have shown high transesterification activity in mild reac- ume (cm3 g1). The triplicate experiments were performed.
tion condition with high stability, non-toxic and low cost [13,14]. The Hammett indicators used were phenolphthalein (pKa = 9.3),
CaO is generally prepared from calcium carbonate (CaCO3) or cal- thymolphthalein (pKa = 9.9), indigo carmine (pKa = 12.2), 2,4-dini-
cium hydroxide (Ca(OH)2) as raw materials. Moreover, using natu- troaniline (pKa = 15) and 4-nitroaniline (pKa = 18.4). Typically, 1–2
ral calcium sources as raw materials for generating CaO could pieces of the catalyst were mixed with 5 mL of a solution of Ham-
eliminate the wastes and at the same time produce a catalyst with mett indicators in methanol and allowed to sit for at least 10 min.
high cost effectiveness [15]. CaO powder prepared from several After the equilibration, the color of the catalyst was noted. The ba-
natural calcium wastes has been used in transesterification of oils sic strength of the catalyst was taken to be higher than the weakest
excluding coral [15–18]. Waste coral fragments are a natural cal- indicator that underwent a color change and lower than the stron-
cium source and is a highly porous material. The use of CaO derived gest indicator that underwent no color change.
from a coral fragment in a bulk solid form could be an effective
way for practical, low cost industrial biodiesel production. Herein,
we present an environmentally friendly, efficient, simple transe- 2.3. Transesterification procedure
sterification process using calcined waste coral fragments as a bulk
solid catalyst. Several parameters are investigated including calci- The transesterification was carried out in a batch reactor. A mix-
nation time of the catalyst, catalyst/oil ratio, methanol/oil molar ture of methanol and catalyst at a designated amount in a 250 ml
ratio, type of oil and FAA quantity, and reuse and recycle of the cat- 2-neck round bottom flask equipped with a reflux condenser was
alyst. The fuel properties of the biodiesel and purity of glycerol are heated at 65 °C for 2 h, and then a preheated 50 ml of oil was
evaluated using standard methods. added. The reaction mixture was heated for the required reaction
times without stir unless mentioned. To follow the reaction pro-
gress, the solution mixture of 0.5 ml was sampled, and then exces-
2. Experimental sive amount of methanol was evaporated in an oven before
analysis of biodiesel yield. The conversion of the oil to methyl es-
2.1. Materials and catalyst preparation ters was determined in term of %FAME as a function of time by nu-
clear magnetic resonance (NMR). The biodiesel yield was
Waste coral fragments collected from Andaman Sea in Krabi calculated from the ratio of the integrations of the 1H NMR signals
province, Thailand, were cut into small pieces (0.5–1.0 cm3) of the methoxy protons (CH3OA) in the methyl ester at chemical
(Fig. 1A), cleaned with water several times and air-dried for 24 h. shifts of 3.68 ppm (singlet peak) and methylene protons (ACH2A)
The coral fragment catalyst was then prepared by calcination at of all triglycerides at chemical shifts of 2.30 ppm (triplet peak),
700 °C in air (refer to TG/DTA results) with calcination times of according to the following equation [21]:
0.5, 1.0 and 1.5 h. Palm oil (FFA of 0.28 mg KOH g1), soybean oil
(FFA of 0.18 mg KOH g1), rice bran oil (FFA of 0.1 mg KOH g1)  
2ACH3
and waste cooking oil (FFA of 1.15 mg KOH g1) were obtained %FAME ¼  100
3ACH2
from commercial sources in Thai market. FFAs (free fatty acids)
content of oil was determined by reported method [19]. The ana-
ACH3 is the integration value of the methoxy protons and ACH2 is the
lytical grade methanol purchased from Fluka was used in this
integration value of the methylene protons. The factors 3 and 2 de-
work. Hammett indicators such as phenolphthalein, thymolphtha-
rive from the number of attached protons at the methoxy and
lein, indigo carmine, 2,4-dinitroaniline, and 4-nitroaniline of AR
methylene carbons, respectively.
grade were obtained from Aldrich and Fluka. DowexÒ 50WX8
It should be noted that the triplicate experiments were per-
ion-exchange resin was obtained from Acros.
formed, and the errors of %FAME contents were typically within
±0.5 %. After completion of the reaction, the solution mixture
2.2. Material characterization was simply separated from a catalytic solid by pouring off from
the reaction vessel. The liquid product obtained was heated in an
Themogravimetric analysis (TG) was carried out on a Rigaku TG- oven (60–70 °C) or rotary evaporated to remove the excess meth-
DTA 8120 thermal analyzer under air flow condition with a temper- anol, and then settled in a separating funnel. The upper phase con-
ature ramp rate of 10 °C/min. X-ray diffraction (XRD) was analyzed sisted of biodiesel product (FAME) and the lower phase contained
by a PHILIPS X’Pert-MDP X-ray diffractometer using Cu Ka radiation the glycerol by-product. Biodiesel was filtered through a cotton
(k = 1.5418 Å) at 1400 W, 40 kV and 35 mA in the scanning angle wool filter and corrected without any treatment and cleaning
(2h) of 20–80° with resolution of 0.04° at counting step of 1 s/step. processes.
X-ray fluorescence was determined by a PHILIPS MagiX wavelength In order to examine the potential reutilization of the catalyst,
dispersive X-ray Fluorescence (XRF) spectrophotometer with 1 kW reuse and recycling of used catalyst was tested to check its capacity
Rh Ka radiation. Fourier transforms infrared (FTIR) spectra were to provide the same catalytic activity. For reuse experiment, the so-
measured on a Perkin–Elmer FTIR spectroscopy spectrum RXI spec- lid catalyst recovered by filtration was directly used as the catalyst
trometer in the range of 400–4000 cm1 with resolution of 4 cm1. for the repeated reactions. The repeated reactions were run under
Potassium bromide (KBr) was used as a matrix. 1H nuclear magnetic the same reaction conditions to the last reaction. For recycling
resonance (1H NMR) spectra were recorded on a Brüker AVANCE experiments, deactivated catalyst was regenerated by washing
300 MHz spectrometer with tetramethylsilane (TMS) as the internal with acetone to remove all remaining reactants, drying at 110 °C
reference using CDCl3 as a solvent in all cases. in an oven and calcining at 700 °C for 1 h. Furthermore, the leach-
Pore volume of the catalyst was measured by water adsorption ing of the catalyst was also tested by removing the solid catalysts
method [20]. Typically, calcined coral fragment preheated in an from the reaction solution after reaction time of 2 h, and then the
oven at 110 °C for 24 h with accurate weight (0.6–0.8 g) was boiled reaction (a remained liquid mixture) was allowed to progress un-
in distilled water (100 ml) in a 250 ml beaker for 2 h. The beaker der the same conditions.
196 W. Roschat et al. / Fuel 98 (2012) 194–202

(B) -6
(A) 100

90
-4

Weight %
80

DTA (µg/s)
70
-2

Uncalcined Calcined 60
coral fragment coral fragment

50 0

100 200 300 400 500 600 700 800

o
Temperature ( C)
(C)
(D)
Calcined coral
Calcined coral, 1.5 h 523 Ca-O

Calcined coral, 1.0 h Uncalcined coral

Transmission (a.u.)
Intensity (a.u.)

Calcined coral, 0.5 h

Ca(OH)2

Uncalcined coral
3637 O-H

Ca(OH)2 529 Ca-O

CaO

CaO
CaCO3
calcite CaCO3 (calcite) 2920 C=O 1787 C=O
2519 C=O 1480 C=O 713 C=O
856 C=O
20 30 40 50 60 70 80 4000 3500 3000 2500 2000 1500 1000 500
2 Theta (degree) Wavenumber (cm-1)

Fig. 1. (A) Pictures of uncalcined and calcined coral fragments. (B) TG/DTA thermograms of coral fragments. (C) Powder XRD patterns and (D) IR spectra of CaCO3, CaO,
Ca(OH)2 and coral fragments calcined at 700 °C in open air with calcination times of 0 (pristine), 0.5, 1.0 and 1.5 h.

To determine the acceptability of this biodiesel product, it was the view point of preparation time and energy consumption, the
tested for standard fuel properties, which are methyl ester content, temperature of 700 °C was selected as a suitable condition to gen-
viscosity, flash point, acid number, density, water content, total erate the catalyst while the calcination time was varied. The coral
contamination and copper strip corrosion, by the American Society fragments, after being calcined at 700 °C for 0.5, 1.0, and 1.5 h,
for Testing and Material (ASTM) methods. Purity of glycerol by- showed porous structure like the uncalcined (pristine) sample
product was analyzed according to the International Standards (Fig. 1A). A high porosity catalyst is a key requirement to achieve
Organizing (ISO) and British Standards Institution (BS) methods. high conversion efficiency for heterogeneous process, thereby high
Analysis of leaching calcium ion from the catalyst in the obtained surface areas or high catalytic sites are necessary. As this catalyst
biodiesel and glycerol was performed by the atomic absorption was a microporous material, its surface area was determined as
spectrophotometric (AAS) method according to the report proce- the total pore volume by water adsorption method, which was
dure [22]. 0.172 ± 0.012 cm3 g1. As illustrated in Fig. 1C, XRD patterns of
Removal of the soluble calcium ion in the obtained biodiesel all the calcined coral fragments were identical, and displayed clear
was performed using commercially available cation-exchange res- and sharp peaks agreeing to a single crystalline phase of CaO. Tiny
ins. The resin was repeatedly immersed in methanol to remove XRD peaks identical to incompletely calcined CaCO3 were detected
moisture, and then was packed into a chromatography column. in the sample with short calcination time of 30 min. Weak XRD
The oil was funneled into the column at room temperature. peaks matching to Ca(OH)2 patterns were observed in all calcined
samples, which might arise from a hydration of CaO by moisture
during storage of the samples. This phenomenon has been ob-
3. Results and discussion served in other forms of CaO and it is intense in case of powder
CaO [23]. It should be noted that the activated CaO was rapidly hy-
3.1. Catalyst preparation and characterization drated, and hence in order to prevent deterioration of the catalyst
reaction must be taken with extreme care to avoid any contact of
TG/DTA analysis of the coral fragment revealed a 44% weight the CaO with room air. The diffraction pattern of the uncalcined
loss at temperatures between 600–740 °C, where carbon dioxide sample showed sharp peaks agreeing to calcite CaCO3. Its IR
(CO2) was removed from calcium carbonate (CaCO3) to give CaO spectrum also showed absorption bands of the CO2 3 group which
(Fig. 1B). Therefore, high temperature above 600 °C was required were identical to those of the CaCO3 (Fig. 1D). The IR spectra of cal-
to transform CaCO3 (pristine coral) to CaO (catalytic coral). From cined samples revealed an absorption band corresponding to the
 W. Roschat et al. / Fuel 98 (2012) 194–202 197

stretching mode of the CaO at 523 cm1 with no absorption bands the optimum load of the catalyst in the present study is 100 wt.%
of CO23 group being observed. These results indicated that all (50 g of the catalyst to 50 g of oil).
CaCO3 in the samples (calcined for 1 and 1.5 h) was completely From the view point of preparation time and energy consump-
transformed to CaO, which can be hydrolyzed by moisture to tion, the catalyst calcined for 30 min was employed. The result re-
Ca(OH)2 by a prolonged exposure to open air. Elemental composi- vealed a clear decline in catalytic activity and the reaction was
tion analysis of all calcined samples with XRF spectroscopy indi- completed (FAME >95%) in a longer time of 7 h compared with
cated that they were composed of mainly CaO 91.9 wt.%, SiO2 the catalyst calcined for 1 h (Fig. 3). This might be because the
2.0 wt.%, SO3 0.6 wt.%, SrO 1.9 wt.% and other oxides 3.6 wt.%. sample calcined at 700 °C for 30 min has a lower amount of cata-
The strengths of the basic sites of all calcined coral fragments mea- lytic CaO as mentioned earlier. Moreover, this could be clearly seen
sured by the Hammett indicator method were found to possess H_ from the reaction catalyzed by uncalcined catalyst, showing little
values in the range of 13.0–15.0, as evidenced by their ability to ef- to no catalytic activity for transesterification. Accordingly, the
fect a color change with 2,4-dinitroaniline (pKa = 15), while the H_ reaction using catalyst calcined 700 °C for 1.5 h was performed
value of uncalcined coral fragment was less than 8.0–10.0, as no with an expectation of improvement in catalytic activity and reac-
color change was observed with phenolphthalein (pKa = 9.3). tion time due to the CaO catalyst should be completely formed.
Unfortunately, a slight decline of the catalyst activity was observed
3.2. Catalyst properties in transesterification and FAME exceeding 95% was reached in 5 h. This would be due to
a prolonged calcination process which brought about the sintering
In heterogeneous transesterification of oil (mixed triglyceride) of the catalyst and thus the lowering of the surface area. The
with methanol catalyzed by CaO, such triglyceride was transesteri- shrinkage of the catalyst grains would eventually suppress the ac-
fied to di- and mono-glycerides, subsequently to FAME (biodiesel). tive sites. Therefore, the optimum calcination condition in the
The basic site of the CaO catalyst is generally known to be active spe- present study is at temperature of 700 °C for 1 h. However, the
cies in catalytic process and the amount of the basic sites should de- higher temperature is not preferable due to the severe sintering
pend on the surface area. During the reaction, the calcium rate and the high energy consumption.
methoxide species, Ca(OCH3)2, was reported to form on the surface The influence of the reaction conditions on the transesterifica-
of the CaO catalyst, and enhanced the transesterification reaction tion of palm oil, such as methanol/oil molar ratio and stirring the
between triglyceride and methanol. Due to the high surface area reaction, was investigated. The methanol/oil ratio is one of the
of powder CaO catalyst, around 3–10 wt.% of the catalyst is gener- most important factors affecting the yield of biodiesel. Although
ally required for a complete transesterification of oil. In our case, stoichiometric ratio requires three moles of methanol for each
the CaO catalyst is in bulk solid form (size of 0.5–1.0 cm3) having mole of oil, the transesterification is commonly carried out with
a surface area significantly lower than that of the powder, thus high- an extra amount of methanol in order to shift the equilibrium to-
er catalyst loading is required. As the loading amount of the catalyst wards the direction of methyl ester formation. According to Centi-
was increased from 70 wt.% to 100 wt.%, the reaction time required kaya et al., for example, transesterification was insufficient at the
to complete a transformation of all triglycerides to biodiesel with molar ratios of methanol/oils below 5:1 [24]. Additionally, the
FAME exceeding 96% decreased (Fig. 2). The reaction with methanol/oil ratio is associated with operating parameters such
100 wt.% of the catalyst reached a completion in 4 h. The result as the type of catalyst used and the quality of oils. The effect of
was basically related to the amount of the catalytic surface area. methanol/oil molar ratio on the transesterification was investi-
However, when the loading amount of the catalyst was higher than gated and found that the reaction advanced faster by increasing
100 wt.%, the reaction time required to complete a reaction in- the methanol/oil molar ratio from 12:1 to 15:1 (Fig. 4). In fact
creased. It was noticed that from this point the amount of the cata- the higher amount of methanol shifted the reaction equilibrium
lyst exceeded the amount oil and methanol. This might lead to the to the forward direction by promoting the formation of methoxide
reactants (methanol vapor) trapped inside micropores of the cata- species on CaO surface. However, further increase in methanol/oil
lyst resulting in a decrease in the amount of methanol available molar ratio up to 18:1 and 21:1 did not promote the reaction. It
for the reaction, thereby lower conversion efficiency. Therefore, was considered that the glycerol would largely dissolve in exces-
sive methanol and subsequently inhibited the reaction of methanol

100
100

80
80

60
%FAME

60
%FAME

Calcination time
1.5 h
Amount of the catalyst
40 1.0 h
(wt% catalyst/oil) 40 0.5 h
70% 0 h (Uncalcined)
80%
20 100%
20
130%
150%

0 0
0 4 8 12 16 20 24
Time (h) 0 4 8 12 16 20 24
Time (h)
Fig. 2. Effects of loading amount of the catalyst on the %FAME: palm oil, catalyst
(calcined at 700 °C for 1 h); methanol/oil molar ratio (15:1); reaction temperature Fig. 3. Effects of calcination time on the %FAME: catalyst (100 wt.%, calcined at
(65 °C) without stir. 700 °C); methanol/oil molar ratio (15:1); reaction temperature (65 °C) without stir.
198 W. Roschat et al. / Fuel 98 (2012) 194–202

100
100

80
80

60

%FAME
60
%FAME

Rice bran oil

(FFA 0.10 mg KOH/g(oil))
40 40 Soybean oil
(FFA 0.18 mg KOH/g(oil))
Palm oil
Methano/oil ratio (FFA 0.28 mg KOH/g(oil))
20 12:1 20
Waste cooking oil
15:1 (FFA 1.15 mg KOH/g(oil))
18:1
0 21:1 0

0 2 4 6 8 10 0 2 4 6 8 10 12
Time (h) Time (h)

Fig. 4. Effects of methanol/oil molar ratio on the %FAME: catalyst (100 wt.%, Fig. 6. Effects of type of oil and FAA quantity in oil on the %FAME: catalyst
calcined at 700 °C for 1 h); reaction temperature (65 °C) without stir. (100 wt.%, calcined at 700 °C for 1 h); methanol/oil molar ratio (15:1); reaction
temperature (65 °C) with stir.

with oil and catalyst. The optimum methanol/oil molar ratio in the
present reaction is 15:1, which is more than double of the practical 2 h. Moreover, no soap formation was observed in the transesteri-
methanol/oil molar ratio (6:1) for homogeneous process. fication of waste cooking oil (high FFA oil). It was noticed that the
The effect of stirring the reaction mixture was also investigated. initial rates of the reaction were reduced as the amount of FFA in
The separation phase between hydrophilic methanol and hydro- oil increased. At a reaction time of 1 h, FAME was formed in
phobic oil and between both liquid reactants and solid catalyst descending order as follows: rice bran oil (42.6%) > soybean oil
are generally known to be major problems of lowering a conver- (36.4%) >palm oil (23.0%) > waste cooking oil (16.7%). In spite of
sion of reaction in the heterogeneous process. Therefore, combin- the transesterification activity of calcined coral fragment is not as
ing the reaction mixture by stirring would promote the efficient as that of CaO powder and KOH catalysts. However, this
transesterification. As depicted in Fig. 5, the reaction with a con- can be compensated because the coral fragment-based CaO has a
stant stirring progressed at a faster rate and FAME yield over 98% major advantage of easy separation of biodiesel product from the
was accomplished within 2 h, which was only 30 min slower than reaction mixture. Moreover, this catalyst presents good catalytic
that of the typical heterogeneous transesterification of palm oil activity in a bulk solid form, opening the possibility to perform a
catalyzed by CaO powder [22]. continuous catalytic process for biodiesel production in both slurry
To investigate a versatile catalytic activity of the coral fragment and fixed bed configuration reactors, with the capability to use
catalyst, many kinds of oils including palm oil (FFA of 0.28 mg KOH cooking oil as an inexpensive feedstock, which will contribute to
g1), soybean oil (FFA of 0.18 mg KOH g1), rice bran oil (FFA of the elimination of a residue together with the production of
0.1 mg KOH g1) and waste cooking oil (FFA of 1.15 mg KOH g1) biodiesel.
were used in biodiesel synthesis. It was found that the source of
oils (triglyceride) and the amount of free fatty acid (FFA) in the 3.3. Reuse, recycle and stability of catalyst
oil had no effect on the conversion efficiency of the catalyst
(Fig. 6). In all cases, FAME exceeding 98% was achieved within The recycling of the catalyst is an important factor for the eco-
nomical application of coral fragments as heterogeneous catalyst
for biodiesel production. It is also vital to ensure that the active
species are not leached from the solid catalyst. The reusability
100 was checked by using the catalyst without any further activation.
The results (Fig. 7) indicated that the coral fragment catalyst with-
out any further activation could be reused up to five times and
80 more than 94% of FAME reached within 2 h with no obvious loss
of activity. After that the activity of this catalyst was gradually re-
60
duced. The catalyst was completely deactivated after being used
%FAME

more than 10 times. The deactivation of this catalyst might be

due to damage of the active sites on its surface. This damage could
40 With stir result from impurities or from the reactions or the hydration of
Without stir CaO to form Ca(OH)2 which then undergoes carbonation to CaCO3
over the time, because the reactants contain a little amount of
20
water and the reaction was carried out in an open air. The IR and
XRD patterns of the deactivated catalyst showed peaks consistent
0 to CaCO3 as illustrated in Fig. 8. These results were in agreement
with the data reported by Granados et al. [23], and Lanas and
0 2 4 6 8 10 12 Álvarez [25] who also found that CaCO3 developed at the expense
Time (h) of Ca(OH)2. This indicated that hydration of an activated CaO might
Fig. 5. Effects of stirring reaction on the %FAME: palm oil, catalyst (100 wt.%,
not be the main deactivating process as in fact the surface state of
calcined at 700 °C for 1 h); methanol/oil molar ratio (15:1); reaction temperature the working catalyst must be fully hydroxylated but the carbon-
(65 °C). ation of the Ca(OH)2 formed completely eradicated the catalyst.
 W. Roschat et al. / Fuel 98 (2012) 194–202 199

100 another preheated flask, then the reaction was allowed to proceed
New catalyst under the same conditions and the %FAME was monitored. As illus-
1st Recycle catalyst trated in Fig. 9, it was found that no significant increased in %FAME
80 after the catalyst was removed from the reaction. The increase of
%FAME in the early stage after removing of the catalyst, which
reached a constant yield of 30% after 30 min, might come from
60
the reaction of remaining methoxide anion formed during the
%FAME

presence of the catalyst with the triglycerides. However, a slight

40 dissolution of CaO in the reaction media or the leaching of active
species from the surface of the solids cannot be fully discarded
since some research groups have pointed out that slight amounts
20 of CaO were dissolved and transformed into calcium diglyceroxide
by combining with a glycerol by-product during transesterification
[23,27,28]. Kouzu et al. has mentioned that the transesterification
0
was catalyzed not only on basic sites on the surface of the CaO cat-
0 2 4 6 8 10 alyst, but also by the soluble substance [22]. Therefore, the amount
Reused time of soluble substances leaching away from the catalyst was esti-
mated from ions contained in biodiesel and glycerol by-product
Fig. 7. Effects of reuse and recycle of the catalyst on the %FAME: catalyst (100 wt.%,
and compared with that of CaO powder. As shown in Table 1, cal-
calcined at 700 °C for 1 h); methanol/oil molar ratio (15:1); reaction temperature
(65 °C) with stir. cium contents in the biodiesel and glycerol obtained without any
treatment and cleaning processes were 109 and 904 ppm, respec-
tively, for the coral fragment, and 309 and 2439 ppm, respectively,
(A) for the CaO powder. The lower soluble ions content in the products
Absorption intensity (a.u.)

Deactivated coral
from coral fragment catalyst can be rationalized by a bulk solid
form of the catalyst. This indicates that it is more stable to the
Uncalcined coral leaching process than other forms. The calcium content in biodiesel
from both catalysts was somewhat higher than the EN14124 stan-
dard that limits the calcium content in biodiesel to 5 ppm. How-
CaCO3 (calcite) ever, removal of the soluble substance of this magnitude from
the biodiesel can be achieved by cation-exchange resin. Kouzu et
al. also demonstrated that the sulfonated resin efficiently removed
calcium ion from biodiesel with the estimated purification ability
4000 3500 3000 2500 2000 1500 1000 500
of 400 ml oil per 1 ml resin [22]. Indeed, the calcium content of
Wavenumber (cm-1)
the obtained biodiesel that passed through a column packed with
Dowex cation-exchange resin was close to 0 ppm (purification
(B) ability of 200 ml oil/ml resin) (Table 1). Therefore, with the help
of this purification technique, calcined coral fragments can be uti-
Intensity (a.u.)

Uncalcined coral
lized as efficient heterogeneous catalysts for biodiesel production.
Deactivated coral
3.4. Biodiesel and glycerol properties

CaCO3 In the view point of using this coral fragment catalyst in practi-
cal industrial biodiesel production, a primary investigation on
20 30 40 50 60 70 80
100
2 Theta (degree)

Fig. 8. (A) IR spectra and (B) XRD patterns of the deactivated catalyst compared
with CaCO3. 80

However, the catalyst could be regenerated just by simple calcina- 60

%FAME

tion. After activation, the regenerated coral fragment could be re-

used for three times with more than 95% of FAME reached
within 2 h. Therefore, coral fragment catalysts can be reused and 40
regenerated. Many research groups have also recently demon-
strated that the CaO in powder forms can be reutilized several
times [26]. These results stress the possibility of using CaO in bulk 20
solid form as a catalyst when compared with CaO powder and KOH
dissolved in methanol.
Since some calcium ions were found in biodiesel obtained from 0
in

in

in

in

in

in

heterogeneous process catalyzed by powder CaO [22,27] and in

m

m
0

10

30

60

90

general CaO is slightly soluble in methanol (0.035 wt.%) [28], this

12

factor was taken into consideration in our case. In an attempt to Reaction time (after removing the catalyst) (min)
clarify this point the quenching and leaching tests were performed Fig. 9. Effects of reaction time after removing the catalyst at 30 min on the %FAME:
in which the reaction was carried out in the same manner. After a catalyst (100 wt.%, calcined at 700 °C for 1 h); methanol/oil molar ratio (15:1);
30 min reaction, a liquid mixture was transferred carefully to reaction temperature (65 °C) with stir.
200 W. Roschat et al. / Fuel 98 (2012) 194–202

Table 1
Fuel properties of biodiesel obtained from the upper phase of the separation without any treatment and cleaning process.

Fuel propertiesa Standardk Trakarnpruk [29]/Gao [30]/Baroutian [11] This worko

b
Density 860–900 877/–/877 874
Kinematic viscosityc 3.5–5.0 4.5/4.37/5.16 4.33
Flash pointd >120 142/150/188 238
Acid numbere <0.5 0.20/0.3/0.33 0.25
Water contentf <0.050 –/0.039/– 0.025
Copper strip corrosion Number 1 –/–/– Number 1
Methyl ester contentg 96.5 97.9/99/– 96.8
Oxidation numberh >6 –/–/– 12.68
Total contaminationi <24 –/–/– 7.0 ± 0.6
Calcium contentj 5 309 ± 0.1m 109 ± 0.3 (1.1 ± 0.5)n
a
Preformed by biodiesel testing unit, the National Science and Technology Development Agency, Thailand.
b
@15 °C (kg m3).
c
@ 40 °C (cSt).
d
°C.
e
mg KOH g1 oil.
f
%w/w oil.
g
%.
h
hour.
i
mg/kg (ppm).
j
ppm g1 oil.
k
EN 14214 standard.
m
Using CaO powder as catalyst.
n
After purification by cation-exchange resin.
o
Palm oil.

medium scale biodiesel production from palm oil (5 L) was carried prepared from palm oil by other investigators [11,29,30]. The re-
out using small scale optimum conditions: catalyst calcined at sults are summarized in Table 1. The biodiesel product had a den-
700 °C for 1 h; catalyst/oil ratio of 100 wt.%; methanol/oil molar ra- sity of 0.874 g/ml which was slightly higher than that of diesel fuel
tio of 15:1; reaction temperature of 65 °C without stirring. The (0.83 g/ml). According to EN 14214, the minimum acceptable pur-
transesterification reaction progressed slower than a small scale ity for biodiesel is 96.5% of methyl esters. The biodiesel produced
reaction (50 ml) and a complete conversion with FAME of 98% had somewhat higher methyl ester content of 96.8%. The flash
was attained within 8 h, which was twice longer than that of a nor- point in this study was 238 °C, which was much higher than that
mal small scale reaction (Fig. 10). However, we considered that of the compared results and biodiesel obtained from other oils.
with optimum reaction conditions and with stirring, the transeste- The water content in biodiesel product was 0.025 %w/w, which
rification of oil to biodiesel would proceed faster. It is essential to was lower than the limit of standard diesel and biodiesels prepared
note that the biodiesel product and glycerol by-product were cor- by other heterogeneous catalysts. The viscosity of the obtained bio-
rected by simply pouring off from the reaction vessel followed by diesel was 4.33 cSt, which was slightly lower than that of the com-
filtering through a cotton wool filter without any treatment and pared results. The acid value in this study was 0.25 mg KOH/g
cleaning processes. Any excess methanol was rotary evaporated, indicating less content of free fatty acid in the biodiesel product.
and biodiesel product and glycerol by-product were allowed to iso- The total contamination defined as the insoluble material retained
late in separating funnels. Fuel properties of biodiesel were evalu- after the filtration of a heated biodiesel over a standardized 0.8 lm
ated according to the ASTM methods and compared with biodiesels filter was 7.0 ± 0.6 mg/kg, which was fairly lower than the limit of
standard requirement. Referring to EN 14214 standards for bio-
auto fuels, the biodiesel obtained from this study was found to
be within the limits in terms of methyl ester content, density, vis-
100 cosity, flash point, acid value and water content, and in accordance
with other investigators.
80 The properties of the by-product of this biodiesel production
without any treatment and cleaning process was briefly evaluated
according to the International Standards Organizing (ISO) and Brit-
60
%FAME

ish Standards Institution (BS) methods indicating good-quality

crude glycerol compared to crude glycerol obtained from CaO pow-
40 der and KOH catalysts (Table 2, Fig. 11). The attained glycerol was
clear viscous with odorless and color body of light pale. The UV–Vis
spectra in transmission mode displayed percent transmission (%T)
20
over 97% in visible region (400–800 nm). The IR spectra were iden-
tical to that of the 99.5% pure glycerol sample. As a result, to purify
0 this glycerol to higher grades, chemically pure (CP) and United
States Pharmacopeia (USP), might be simply accomplished by
0 2 4 6 8 10 extraction with hexane, ion removal by activated carbon adsorp-
Time (h) tion followed by distillation, thereby high-value glycerol. Full
investigations are in progress. The value of crude glycerol in the
Fig. 10. Effects of reaction time on the %FAME: 5L scale reaction, catalyst (100 wt.%,
calcined at 700 °C for 1 h); methanol/oil molar ratio (15:1); reaction temperature marketplace generally depends on its quality as determined by
(65 °C) without stir. the ease of its subsequent refining.
 W. Roschat et al. / Fuel 98 (2012) 194–202 201

Table 2
Properties of glycerol obtained from the lower phase of the separation with further heated in an oven (70–80 °C) for 24 h followed by filtration through a cotton wool filter.

Properties Standard Catalyst/oil

Coral fragment/palmd Coral fragment/sunflowerd CaO powder/palme KOH/palme
a
Glycerol content (%) 95 68 78 44 23
Total ash (%) 10b 0.302 0.022 0.12 5.38
Sulfated ash (%) 0.01a 0.026 0.010 0.45 4.22
pH – 7.32 7.72 7.82 10.34
Density (g cm3) 1.26c 1.18 1.20 1.10 <1.00
Calcium content (ppm/g) 5a 904 ± 0.7 – 2439 ± 0.9 –
Color Water-whitec Light pale Light pale Light pale Brown
a
Pharmaceutical grade.
b
Crude glycerol from soap manufacture.
c
Pure glycerol.
d
Using optimum conditions.
e
Using literature methods [5,16,17].

Coral/sunflower oil
(A) 100 (B)
Coral/palm oil
80
Pure glycerol
%Transmittance

%Transmittance
60 CaO powder/palm oil

CaO powder/palm oil

KOH/palm oil
40 KOH/palm oil
Pure glycerol
Coral/palm oil
Coral/sunflower oil
20

0
200 300 400 500 600 700 800 4000 3500 3000 2500 2000 1500 1000 500
Wavelength (nm) Wavenumber (cm-1)

Fig. 11. UV–Vis and IR spectra of by-product glycerol obtained without any treatment and cleaning processes.

4. Conclusion Acknowledgements

In summary, we have demonstrated that solid catalysts derived This work was supported by the Thailand Research Fund
from calcination of waste coral fragments can be utilized as effi- (IUG5180010), Strategic Scholarships for Frontier Research Net-
cient heterogeneous solid catalysts for transesterification of many work for Research Groups (CHE-RES-RG50) funded by the Office
kinds of oils to biodiesel. The reaction under optimum conditions of the Higher Education Commission, Thailand. We acknowledge
of catalyst calcined at 700 °C for 1.0 h; catalyst/oil ratio of the scholarship support from the Thailand Graduate Institute of
100 wt.%; methanol/oil molar ratio of 15:1; reaction temperature Science and Technology (TGIST) to Mr. Wuttichai Roshchat and
of 65 °C with a constant stirring yields biodiesel with a FAME Center for Innovation in Chemistry (PERCH-CIC).
exceeding 98% in 2 h. The catalyst can be easily separated from
the reaction mixture by simple filtration and can be used several
times with consistent activity. Biodiesel and glycerol products
without any treatment and cleaning processes show high quality References
with fuel properties of the biodiesel meet all EN 14214 standard.
[1] Sharmaa YC, Singha B, Upadhyayb SN. Advancements in development and
The calcined coral fragment presents good catalytic activity in a characterization of biodiesel: a review. Fuel 2008;87:2355–73.
bulk solid form, opening the possibility to perform a continuous [2] Lam MK, Lee KT, Mohamed AR. Homogeneous, heterogeneous and enzymatic
catalytic process for biodiesel production with batch reactor. The catalysis for transesterification of high free fatty acid oil (waste cooking oil) to
biodiesel: a review. Biotechnol Adv 2010;28:500–18.
utilization of this good-quality glycerol together with the natural [3] Shahid EM, Jamal Y. A review of biodiesel as vehicular fuel. Sust Energy Rev
availability of the coral fragment and the capability to use a residue 2008;12:2484–94.
as cooking oil as inexpensive feedstock would supplementary im- [4] Phan AN, Phan TM. Biodiesel production from waste cooking oils. Fuel
2008;87:3490–6.
prove the economic aspect of overall biodiesel production based
[5] Li Y, Qiu F, Yang D, Li X, Sun P. Preparation, characterization and application of
on this catalyst, producing cost-efficient biodiesel. The catalyst, heterogeneous solid base catalyst for biodiesel production from soybean oil.
therefore, displays high potential as solid catalyst for practical, Biomass Bioenergy 2011;35:2787–95.
[6] Refaat AA. Biodiesel production using solid metal oxide catalysts. Int J Environ
environmentally friendly, and low cost industrial biodiesel and
Sci Tech 2011;8:203–21.
glycerol production. For people in rural areas or developing coun- [7] Silva CCCM, Ribeiro NFP, Souza MMVM, Aranda DAG. Biodiesel production
tries, this process could be an effective way to obtain cheap energy from soybean oil and methanol using hydrotalcites as catalyst. Fuel Process
sources. Full investigation on an industrial scale of biodiesel and Tech 2010;91:205–10.
[8] Carrero A, Vicente G, Rodríguez R, Linares M, del Peso GL. Hierarchical zeolites
glycerol production based on this catalyst is on-going in our as catalysts for biodiesel production from Nannochloropsis microalga oil. Cat
laboratory. Today 2011;167:148–53.
202 W. Roschat et al. / Fuel 98 (2012) 194–202

[9] Long T, Deng Y, Li G, Gan S, Chen J. Application of N-methylimidizolium [19] Virmani OP, Narula AK. Applied chemistry. 2nd ed. New Age International
functionalized anion exchange resin containing NaOH for production of Publishers; 1995.
biodiesel. Fuel Process Tech 2011;92:1328–32. [20] Harry KG, Johnson A. A non-destructive technique for measuring ceramic
[10] Taufiq-Yap YH, Lee HV, Hussein MZ, Yunus R. Calcium-based mixed oxide porosity using liquid nitrogen. J Archaeological Sci 2004;31:1567–75.
catalysts for methanolysis of Jatropha curcas oil to biodiesel. Biomass [21] Gelbard G, Bres O, Vargas RM, Vielfaure F, Schuchardt UF. 1H nuclear magnetic
Bioenergy 2011;35:827–34. resonance determination of the yield of the transesterification of rapeseed oil
[11] Borges ME, Díaz L, Alvarez-Galván MC, Brito A. High performance with methanol. J Am Oil Chem Soc 1995;72:1239–41.
heterogeneous catalyst for biodiesel production from vegetal and waste oil [22] Kouzu M, Yamanaka S, Hidaka J, Tsunomori M. Heterogeneous catalysis of
at low temperature. Appl Catal B Environ 2011;102:310–5. calcium oxide used for transesterification of soybean oil with refluxing
[12] Baroutian S, Aroua MK, Raman AAA, Sulaiman NMN. Potassium hydroxide methanol. Appl Catal A General 2009;355:94–9.
catalyst supported on palm shell activated carbon for transesterification of [23] Granados ML, Zafra Poves MD, Alonso DM, Mariscal R, Galisteo FC, Moreno-
palm oil. Fuel Process Tech 2010;91:1378–85. Tost R, et al. Biodiesel from sunflower oil by using activated calcium oxide.
[13] Kouzu M, Kasuno T, Tajika M, Sugimoto Y, Yamanaka S, Hidaka J. Calcium oxide Appl Catal B Environ 2001;73:317–26.
as a solid base catalyst for transesterification of soybean oil and its application [24] Cetinkaya M, Karaosmanoglu F. Optimization of base-catalyzed
to biodiesel production. Fuel 2008;87:2798–806. transesterification reaction of used cooking oil. Energy Fuels 2004;18:
[14] Boey PL, Maniam GP, Hamid SA. Performance of calcium oxide as a 1888–95.
heterogeneous catalyst in biodiesel production: a review. Chem Eng J [25] Lanas J, Álvarez JI. Dolomitic limes: evolution of the slaking process under
2011;168:15–22. different conditions. Thermochim Acta 2004;423:1–12.
[15] Ngamcharussrivichai C, Nunthasanti P, Tanachai S, Bunyakiat K. Biodiesel [26] Reddy C, Reddy V, Oshel R, Verkade JG. Room-temperature conversion of
production through transesterification over natural calciums. Fuel Process soybean oil and poultry fat to biodiesel catalyzed by nanocrystalline calcium
Tech 2010;91:1409–15. oxides. Energy Fuels 2006;20:1310–4.
[16] Cho YB, Seo G. High activity of acid-treated quail eggshell catalysts in the [27] Kouzu M, Kasuno T, Tajika M, Yamanaka S, Hidaka J. Active phase of calcium
transesterification of palm oil with methanol. Bioresour Technol 2010;101: oxide used as solid base catalyst for transesterification of soybean oil with
8515–9. refluxing methanol. Appl Catal A Gen 2008;334:357–65.
[17] Yoosuk B, Udomsap P, Puttasawat B. Hydration–dehydration technique for [28] Gryglewicz S. Rapeseed oil methyl esters preparation using heterogeneous
property and activity improvement of calcined natural dolomite in catalysts. Bioresour Technol 1999;70:249–53.
heterogeneous biodiesel production: structural transformation aspect. Appl [29] Trakarnpruk W, Porntangjitlikit S. Palm oil biodiesel synthesized with
Catal A Gen 2011;395:87–94. potassium loaded calcined hydrotalcite and effect of biodiesel blend on
[18] Chakraborty R, Bepari S, Banerjee A. Application of calcined waste fish (Labeo elastomers properties. Renew Energy 2008;33:1558–63.
rohita) scale as low-cost heterogeneous catalyst for biodiesel synthesis. [30] Gao L, Teng G, Xiao G, Wei R. Biodiesel from palm oil via loading KF/Ca–Al
Bioresour Technol 2011;102:3610–8. hydrotalcite catalyst. Biomass Bioenergy 2010;34:1283–8.