Coconut Oil
Coconut Oil
abstract
The commercial success of biodiesels has to date been limited by high production costs of vegetable oil methyl esters.
High feedstock costs are compounded by side reactions such as soap formation during conversion using conventional
catalysts, and the consequent costs of product refining and purification. Recent studies on the heterogeneous catalysis of the
transesterification reaction with commercial ion exchange resins have met with some success; the findings suggest that use
of heterogeneous catalyst improves yield compared to conventional processing with homogeneous acid or alkaline catalyst.
This study investigated the performance of heterogeneous calcium oxide catalyst in the production of coconut methyl ester.
Specifically, the study investigated the effect of temperature, time, excess methanol and catalyst to oil ratio on conversion of
oil in batch reactions as well as the level of trace calcium in the final product using a two-level factorial experimental design.
The tests achieved conversion levels of 91.5 – 95.7%, based on measured TG levels of 0.6 – 1.2%. The specific gravity of the
biodiesel phase was also found to be in the range of 0.83 – 0.87, which is also indicative of high conversion. Only temperature
was found to have a statistically significant effect on triglyceride conversion, which implies that the overall rate of reaction
is controlled by surface reaction kinetics rather than mass transfer. On the other hand, none of the experimental factors were
found to have a statistically significant effect on the level of calcium contamination of the biodiesel product.
1.0 INTRODUCTION oil have the largest potential, while other vegetable oils such as
Many countries have started to implement measures to corn and coconut oil are available in smaller quantities. Non-
reduce adverse impacts on their economies through energy traditional feedstocks such as Jatropha curcas and oil-bearing
efficiency and alternative fuel programs. The transportation algae are also the subject of much research interest [1-6].
sector has been affected more than other sectors due to its With currently available raw materials and process
heavy dependence on petroleum products. Alternative fuels technology, the main reasons for the high cost of biodiesel are:
such as biodiesel have been proposed as substitutes to reduce
the vulnerability of net oil importers. In addition, such fuels are • Absence of economies of scale in the emerging biofuels
expected to yield significant environmental benefits such as the industry;
reduction of emissions of greenhouse gases and air pollutants. • High feedstock costs;
Potential feedstocks for biodiesel include soya oil, rapeseed oil, • Side-reactions, such as the saponification reaction, leading to
coconut oil, palm oil, and jatropha oil, with the choice being feedstock losses and increased utility demands for product
dependent on the region. refining.
Biodiesel is now available commercially in limited
quantities, but its acceptance has been hindered due to its high Improvements in process technology can potentially address
cost as well as feedstock supply limitations. Global production of the third factor. In principle, vegetable oils can be converted into
vegetable and marine oils in recent years was about 100 million biodiesels that closely approximate the properties of petroleum
t/a [1], which is mostly dedicated to traditional uses. Sharma and diesel by means of transesterification reactions. Theoretical
Singh [2] estimate the combined biodiesel potential of the top details are discussed in the next section. By far the most commonly
ten producing countries in the world at about 40 × 109 l/a, with used means of making transesterification reactions proceed
production costs ranging from US$0.5 – 1.7/l; significantly, they at commercially useful rates is through the use of alkali catalysts
list three Southeast Asian countries among the top ten: Malaysia, such as NaOH [1-5]. Their main drawback is that they react
Indonesia and the Philippines. In terms of volume, soya and palm with naturally occurring free fatty acids to form soaps, which
contaminate the methyl ester product and thus require subsequent
Journal - The Institution of Engineers, Malaysia (Vol. 70, No.4, December 2009) 19
processing to yield a fuel product of acceptable quality. Thus, Hence, each gram of coconut oil is expected to yield one
the use of alkaline catalysts may also require pretreatment of the gram of methyl ester. The glycerol separates spontaneously from
feedstock oil to remove the fatty acids. the biodiesel phase after the reaction. The latter phase contains
Acid catalysts are more suitable for highly acidic feedstocks; the methyl ester as well as unreacted oil, whose molecular
however, acid catalysed transesterification reactions proceed glycerol content can then be used to assess the extent of the
very slowly, and thus require larger, more expensive reactors for reaction (pure methyl ester should be glycerol-free). Majority of
a given rate of production [4]. There is considerable interest in excess methanol present dissolves in the glycerol phase.
the use of highly selective enzymes of biocatalysts; however, the The reactions given in Equations 1 and 2 take place in the
commercial potential of biocatalysis remains limited by the cost presence of a catalyst, or at high pressures and temperatures, as
of the lipase enzymes. Another catalyst-free alternative requires discussed in the previous section. The presence of impurities,
the use of supercritical methanol at temperatures in the range however, leads to undesirable side reactions yielding products
of 200 – 400 °C. Supercritical plants can achieve high product that require further cleaning and refining. For example, free fatty
yields even with low-grade feedstocks, but they entail high initial acids in the feedstock oil react with alkaline catalysts to give soap.
costs and are energy-intensive. The soap must then be removed from the methyl ester product by
a series of water- and energy-intensive washing operations. Such
2.0 Fundamental Aspects of the side reactions impose three process cost penalties: first, a portion
of the raw material is converted to an undesired product; second,
Transesterification Reaction
additional processing effort and cost are incurred to remove the
Vegetable oils consist primarily of triglyceride molecules undesired product; finally, the side reaction consumes some of
with three fatty acid chains linked to a glycerol backbone. the catalyst that is needed for the main process reaction. Thus,
Naturally occurring oils also contain trace impurities, including if a heterogeneous catalytic system can be developed such that
diglycerides, monoglycerides, free fatty acids and other organic these side-reactions are avoided, then substantial cost savings
compounds. Depending on the precautions taken during the may be realized.
production and storage of vegetable oils, these impurities may Glycerol is the main byproduct of the transesterification
be present in significant amounts which require downstream process. Its presence in biodiesel is highly undesirable as it can
process interventions. For example, for edible oil production, lead to engine damage; on the other hand, there is a considerable
free fatty acids are normally removed by neutralisation with market for glycerol as a chemical commodity in its own right but
NaOH and trace organic compounds with disagreeable flavors or the refined glycerol has a much higher price than the unrefined
colors are removed by adsorption and vacuum steam distillation. version [8]. Thus, one of the main purposes of biodiesel refining
The transesterification reaction for converting triglycerides into is to remove all traces of glycerol from the methyl ester product,
methyl esters is and to recover the glycerol for further use. At the same time,
excess methanol is usually supplied to drive the reaction to
economically viable conversion levels. The unreacted methanol
in the reactor outlet stream must then be recovered and fed back
into the system. The presence of the previously mentioned side
reactions complicate these downstream recovery processes,
and thus the development of more selective reaction pathways
can have indirect benefits on process flow sheets in biodiesel
plants.
This study will provide essential baseline information for the
potential of heterogeneous catalysts in the commercial production
of biodiesel. One of the major biodiesel producers has expressed
(1) its interest in the research; the findings of this study can thus
provide basis to decide on the feasibility of a long-term process
This conversion is necessary if the final product is to be development collaboration with this firm. At the moment, proof
used as engine fuels. Triglyceride molecules are much larger than of concept is necessary before even considering the eventual
hydrocarbon molecules found in diesel fuel; hence, vegetable commercialisation of the technology. In particular, it is essential
oils are more viscous and less volatile than commercial diesel to determine whether it is possible to obtain promising yields and
fuel. Such properties lead to poor combustion and fouling build whether the process is sufficiently robust to deal with feedstocks
up in the fuel injectors of diesel engines. Conversion to methyl of variable quality
ester is necessary to yield a product with smaller molecules, and
properties more closely resembling those of diesel. 3.0 Review of CaO as Heterogeneous
Note that three moles of methanol are needed for each mole
of vegetable oil feed. However, since the reaction is reversible,
Catalysts for Transesterification
excess methanol is normally used to shift the reaction forward. There has been much recent work on the use of various
Furthermore, in the case of coconut oil whose molecular weight heterogeneous catalysts for transesterification, which are
typically is in about 660 g/mole [7], the balanced reaction, summarised in Table 1. In particular, work on the use of
expressed in theoretical mass ratios, is: commercial ion exchange resins has given some promising
results, with conversions of up to 85% being reported for both
Coconut oil + 0.14 Methanol ⇔ 0.14 Glycerine + Methyl Ester pure triglycerides and selected vegetable oils [7, 9-11]. Many of
(2) the studies use soybean oil as feedstock; Marchetti [12] reports
20 Journal - The Institution of Engineers, Malaysia (Vol. 70, No.4, December 2009)
the use of waste frying oils, which contain significant amounts of The study will focus on laboratory scale batchwise
free fatty acids. Conversion levels of about 80% suggest potential processing of coconut oil with methanol using calcium oxide in
for commercial use. Calcium oxide (CaO) catalyst was reported batch reactions. No detailed theoretical investigation of reaction
by Zhu et al. [13] to give conversion levels of up to 93% using mechanism or catalyst characteristics will be done; the emphasis
Jatropha curcas oil as feedstock. Similar conversion levels were was on identifying trends and effects which are useful for future
reported for conversion of soya oil [14]. Significantly they also tests on a larger scale.
found the presence of water to improve catalyst performance.
Repeated use of the catalyst in these tests also showed good 5.0 Methodology
durability. Kouzu et al. [15] tested different calcium compounds
Commercially available technical grade calcium oxide was
as catalysts for conversion of soybean oil, and found CaO to
used as catalyst. The catalyst was pretreated using the procedure
give the best conversion. Subsequent tests in continuous flow
reported by Zhu et al. [13]. The procedure involves immersing
reactors showed some problems with deterioration of catalyst
the catalyst in ammonium carbonate solution (C = 0.12 g/ml).
performance with extended use, as well as contamination of
The catalyst was then filtered, stored in a desiccator for 1 hour,
the biodiesel product with leached calcium [16]. To date, no
screened (mesh size < 60) and then calcined at 900 °C for 1.5 h.
studies have been reported on CaO catalysed transesterification
Finally, the catalyst was cooled down to 250 °C and subsequently
of coconut oil.
stored in a desiccator.
Batch tests were conducted with experimental variables at
Table 1: Some Key Findings on CaO Catalysts for Biodiesel
Production the levels shown in Table 2, based on a 24 factorial experimental
design [18]. These levels were based in part on the experiments
Authors Key Findings of Zhu et al. [13] and also on lessons learned during preliminary
tests. Some of the early tests achieved no discernible conversion,
Kouzu In comparative tests for transesterification of as indicated by the complete failure to form a dense glycerol
et al. soybean oil using calcium oxide hydroxide phase. Furthermore, large amounts of catalyst were found to be
[15, 16] and carbonate, CaO was found to give the counterproductive, due to the entrainment of the calcium oxide
best conversion. Some catalyst deterioration grains in the glycerol byproduct. These preliminary results were
reported in subsequent tests. Contamination of then used to determine the appropriate levels of the experimental
product by calcium leaching also observed variables.
Liu et al. Transesterification of soybean oil using Table 2: Experimental variables and levels
[14] CaO catalyst. Maximum conversion of 95%
Factor Low Level High Level
was reported at 300% excess methanol, 8% Variable
Code ( –1) (+1)
catalyst dose at T = 65 °C and 3 h reaction
time. Presence of water was found to improve A Temperature (°C) 55 65
catalyst performance. B Time (h) 1.5 3
C Excess methanol 100 200%
Lopez Transesterification of soybean oil using
(mole %)
Granados CaO catalyst. Concluded that CaO is easily
et al. [17] deactivated by moisture and CO2 from the D Catalyst-oil ratio 0.005 0.01
ambient air.
The samples were then put in separatory funnels to allow
Zhu et al. Transesterification of Jatropha curcas oil using the biodiesel and glycerol phases to separate. The biodiesel
[13] CaO. Conversion of 93% was reported at phase was then distilled to remove dissolved methanol, and
T = 70 °C, t = 2.5 h, 200% excess methanol and then analyzed for total glycerol (TG) and calcium content. The
1.5% catalyst dose. TG analysis was performed at the fuel testing laboratory of the
Philippine Department of Energy, while calcium analysis was
done in-house using atomic absorption spectrophotometry (AAS)
after sample digestion with nitric acid.
4.0 Objectives
Recent findings described in the previous sections suggest
6.0 Results and Discussion
that the use of heterogeneous transesterification catalysts
for biodiesel production provides a promising alternative to The experimental results of the study are summarised in
conventional processes. It is necessary to conduct tests applied Table 4. The conversion levels shown were computed from the
specifically to the production of CME, in order to determine the TG analysis. As seen in Equations 1 and 2, the amount of TG in
commercial potential of this technology. The objectives of the the biodiesel phase decreases as more of the triglyceride (coconut
study are: oil) is converted into methyl ester. Based on the typical molecular
• To determine the suitability of calcium oxide for use as weight of coconut oil (660 g/mole) the theoretical TG present is
heterogeneous catalysts for the production of CME. 13.9% [7]. On the other hand, pure methyl ester should contain
• To determine the effect of temperature, excess methanol, no bound glycerol. Thus, if the biodiesel phase consists mainly
catalyst/oil ratio, reaction time on triglyceride conversion. of these two components, lower TG values imply higher levels of
• To determine the degree of contamination of the biodiesel methyl ester, and thus, higher conversion. Note that, even in the
product with leached calcium. presence of intermediate species (i.e., mono- and diglycerides)
Journal - The Institution of Engineers, Malaysia (Vol. 70, No.4, December 2009) 21
the TG analysis provides a conservative estimate of conversion. a significant (P-value = 0.025) positive effect of temperature
The mass fraction of methyl ester in the final product may be on conversion, while time (P-value = 0.967), excess methanol
computed from the TG analysis using: (P-value = 0.592) and catalyst-oil ratio (P-value = 0.260) were
found not to have statistically significant effects for the range
TGsample = TGCME XCME + TGoil(1 - XCME) (3) of values used in the experiments. Lower P-values indicate
more significant influence of a given experimental variable on
Where TGsample, TGCME and TGoil are the mass percentages the response; the threshold for a statistically significant effect is
of TG in the sample, in CME and in unconverted coconut oil, typically 0.05 [18]. By comparison, Zhu et al. [13] report 93%
respectively. As noted above, TGCME is 13.9% and TGoil is 0%. conversion of Jatropha curcas oil at comparable experimental
Furthermore, since the theoretical yield for the transesterification conditions. The strong influence of temperature on the conversion
based on Equation 2 is 1 kg of CME for every kg of coconut found in this study contrasts with results reported by Co [7] for
oil, XCME is also equivalent to conversion in the absence of ion exchange resin catalysts. The relatively high sensitivity to
side reactions. For example, if a batch begins with 1 kg of pure temperature (as indicated by the statistically significant effect
coconut oil, 1 kg of CME will be formed once the reaction of temperature on conversion) further suggests that the overall
reaches 100% conversion. Proportionately less CME will form reaction rate is determined by the surface reaction rate rather
at lower conversion levels If, for example, it is found that the than by mass transfer. This result will have implications for
XCME = 0.9, then the product must contain 0.9 kg of CME and scale-up and design of flow reactor systems, which will differ
0.1 kg of coconut oil; hence, this mass fraction indicates that markedly from those found by Co [7], whose findings suggest
90% of the coconut oil initially present has been converted to a mass transfer controlled reaction mechanism. This difference
CME. can be attributed to the smaller particle size of the catalyst used
The conversion levels computed from the TG analysis of in this study. Furthermore, the ion exchange catalyst used by Co
the sixteen batches range from 91.5 – 95.7%. Statistical analysis [7] was in supported form, and thus may have exhibited more
using the multiple regression toolbox of Microsoft Excel showed internal pore resistance.
22 Journal - The Institution of Engineers, Malaysia (Vol. 70, No.4, December 2009)
The biodiesel product from all the tests exhibited turbidity. It high conversion. Statistical analysis of the results of a two-level
is unclear at this point if the turbidity is due to the presence of solid factorial experimental design shows that only temperature has a
calcium oxide particles, droplets of entrained liquid or calcium significant positive effect on the conversion. The other factors
soap. As previous studies have shown calcium contamination of were not found to significantly influence conversion for the
the biodiesel product, the samples from the study were tested range of values investigated here. This statistically significant
for traces of leached calcium using AAS. The calcium content effect of temperature strongly suggests that the overall reaction
of the samples was 28.2 – 135.7 ppm, while a blank biodiesel rate is controlled by surface reaction kinetics, in contrast to
sample (produced using non-calcium based catalyst) was found recent work with ion exchange resins that suggests mass transfer
to contain 42 ppm calcium. Hence, the latter value appears to be controlled reaction rates [7]. The difference may be attributed to
the typical calcium level found in the feedstock. The results are the larger particle size and the supported nature of the catalyst
also shown in the final column of Table 4. Note that multiple used in the previous work. Analysis of the biodiesel samples also
regression analysis showed that none of the experimental factors showed trace contamination with calcium in the range of 28.2
had a statistically significant effect on the amount of calcium – 135.7 ppm. This range is similar to the results reported in the
leached into the biodiesel product; all P-values calculated were literature [15, 16], and is significantly higher than the 42 ppm
larger than 0.05. Also, the level of trace calcium exhibited only level measured in a blank sample consisting of biodiesel made
a weak positive correlation (r = 0.40) with conversion level. using non-calcium based catalyst. However, it is not clear yet
Thus, it is necessary to conduct a more detailed investigation on whether this contamination is in the form of solid particulates
the response of leaching to various process influences in future or ionic calcium in entrained liquid. Furthermore, none of the
studies. experimental variables investigated showed a statistically
significant effect on calcium levels in the biodiesel product. An
7.0 Conclusions extension of this work in the future will be the use of catalyst
in continuous-flow packed bed reactor experiments. More
This study investigated the effect of temperature, time,
detailed investigation of the reaction mechanism, as well as
excess methanol and catalyst-to-oil ratio on the performance of
catalyst robustness and durability, should also be carried out.
pretreated calcium oxide catalyst for the transesterification of
Furthermore, it is essential to test the effectiveness of the catalyst
coconut oil in batch reactions. Performance was measured in
on a wider variety of biodiesel feedstocks.
terms of conversion, which was calculated from the total glycerol
(TG) analysis of the biodiesel phase, and the level of calcium
contamination of the product. The tests achieved conversion Acknowledgment
levels of 91.5 – 95.7%, based on measured TG levels of 0.6 – The authors gratefully acknowledge the financial support
1.2%. The specific gravity of the biodiesel phase was also found of the De La Salle University Research Coordination Office
to be in the range of 0.83 – 0.87, which is also indicative of (URCO) through Project Grant No. 42 B U/C 3 06.
REFERENCES
[1] Demirbas, A. (2007). Importance of biodiesel as a [7] Co, CET (2009). Biodiesel production from coconut oil
transportation fuel. Energy Policy 35: pp. 4661 – 4670. and methanol using heterogenous catalyst in a packed
bed reactor. MSc thesis, De La Salle University, Manila,
[2] Sharma, YC and Singh B (2009). Development of biodiesel: Philippines.
Current scenario. Renewable and Sustainable Energy
Reviews 13: pp. 1646 – 1651. [8] Saka, S and Iwayama Y (2009). A new process for the
catalyst-free production of biodiesel using supercritical
[3] Bondioli, P. (2004). The preparation of fatty acid esters
methyl acetate. Fuel 88: pp. 1307-1313.
by means of catalytic reactions. Topics in Catalysis 27:
pp. 77 – 82.
[9] Lopez, D, Goodwin, JG and Bruce, DA (2007).
[4] Marchetti, JM, Miguel, VU and Erraza, AF(2007). Transesterification of triacetin with methanol on Nafion
Possible methods for biodiesel production. Renewable and acid resins. Journal of Catalysis 245: pp. 379 – 389.
Sustainable Energy Reviews 11: pp. 1300 – 1311.
[10] Shibasaki-Kitakawa, N, Honda, H, Kuribayashi, H, Toda, T,
[5] Basha, SA, Gopal, KR and Jeharaj, S. (2009). A review Fukumura, T and Yonemoto, T (2007). Biodiesel production
on biodiesel production, combustion, emissions and using anionic ion exchange resin as heterogeneous catalyst.
performance. Renewable and Sustainable Energy Reviews Bioresource Technology 98: pp. 416-421.
13: pp. 1628 – 1634.
[6] Razon, L. F. (2009). Alternative crops for biodiesel [11] Xie, W, Huang X and Li, H (2007). Soybean oil methyl
feedstock. CAB Reviews: Perspectives in Agriculture, esters preparation using NaX zeolites loaded with KOH
Veterinary Science, Nutrition and Natural Resources 4: as a heterogeneous catalyst. Bioresource Technology 98:
No. 56. pp. 936 – 939.
Journal - The Institution of Engineers, Malaysia (Vol. 70, No.4, December 2009) 23
[14] Liu X. He, H, Wang Y, Zhu S and Piao X (2008). Byron Bugay
Transesterification of soybean oil to biodiesel using CaO as Student in the BS Chemical Engineering Program of De
La Salle University, Manila. His final year research project
solid base catalyst. Fuel 87: pp. 216 – 221.
focused on the testing of heterogeneous catalysts for biodiesel
production.
[15] Kouzu, M, Kasuno, T, Tajika, M, Sugimoto, Y, Yamanaka, S
and Hidaka, J, (2008). Calcium oxide as a solid base catalyst
for transesterification of soybean oil and its application to
biodiesel production. Fuel 87: pp. 2798 – 2806.
Dr Luis f. Razon
[16] Kouzu, M. Hidaka, J, Konichi, Y, Nakano, H and Yamamoto, Professor of Chemical Engineering at De La Salle University,
M (2009). A process to transesterify vegetable oil with Manila, Philippines. His research interests are chemical
reactor dynamics and the development of novel feedstocks
methanol in the presence of quick lime bit functioning as
and catalysts for biodiesel production. He earned the degrees
solid base catalyst. Fuel 88: pp. 1983:1990. of BS Chemical Engineering (magna cum laude) at De La
Salle University, and MS and PhD Chemical Engineering
[17] Lopez-Granados M, Zafra Poves MD, Martin Alonso D, at the University of Notre Dame, Indiana, USA. He was
Mariscal R, Cabello Galisteo F, Moreno-Tost R, Santamaria later a Postdoctoral Research Associate at the University
J and Fierro JLG (2007). Biodiesel from sunflower oil of Notre Dame, Product Development Manager at Mead
by using activated calcium oxide. Applied Catalysis B: Johnson Nutritionals and Director of the University Research
Coordination Office of De La Salle University. Prof. Razon
Environmental 73: pp. 317-326. is also a member of the National Research Council of the
Philippines (NRCP).
[18] Kiemele, MJ, Schmidt, SR and Berdine, RJ. (2000). Basic
statistics. Tools for continuous improvement. 4th ed. Air
Academy Press, Colorado Springs.
Dr Raymond R. Tan
Professor of Chemical Engineering and University Fellow
at De La Salle University, Manila, Philippines. His research
PROFILEs interests include pinch analysis, life cycle assessment and
process systems engineering. Prof. Tan is the recipient of
Jo Marie Velasquez multiple awards from the National Academy of Science and
Student in the BS Chemical Engineering Program of De Technology (NAST) and the National Research Council
La Salle University, Manila. Her final year research project of the Philippines (NRCP), and currently has over forty
focused on the testing of heterogeneous catalysts for biodiesel published or forthcoming articles in ISI-indexed journals in
production. chemical, environmental and energy engineering. Dr Tan is
also a member of the editorial board of Clean Technologies
and Environmental Policy.
24 Journal - The Institution of Engineers, Malaysia (Vol. 70, No.4, December 2009)