Content uploaded by Wiwat Pichayakorn
Author content
All content in this area was uploaded by Wiwat Pichayakorn
Content may be subject to copyright.
Research Article
Model of Degradation Kinetics for Coconut Oil at
Various Heating Temperatures
Wibul Wongpoowarak, Wiwat Pichayakorn, Kwunchit Oungbho,
Watcharakorn Boontaweesakul, Siriporn Sirivongmongkol
and Prapaporn Boonme*
Department of Pharmaceutical Technology, Faculty of Pharmaceutical Sciences,
Prince of Songkla University, Hat-Yai, Songkhla ,Thailand
*Corresponding author. E-mail address: prapaporn.b@psu.ac.th
Received September 24, 2008; Accepted January 20, 2009
Abstract
This study was aimed to investigate the model of degradation kinetics for coconut oil at various heating
temperatures. The oxidative stability of coconut oil at temperature variations between 130 and 200 °C was
investigated by UV absorbance measurement. The induction times (time-to-abrupt change) were determined
from the curves between the absorbance at 270 nm, related to amount of secondary oxidation products, and
time. The model of degradation kinetics was empirically estimated using non-linear regression. From the
obtained model, coconut oil degraded via at least two different pathways while being heated, i.e., autocatalysis
and first order reactions. This model could describe the entire degradation curve including the induction time
behavior of coconut oil at heating temperatures below 150 °C. However, at higher heating temperatures, the
other pathways of degradation might occur which could not be explained by the proposed model.
Key Words: Degradation kinetics; Oxidative stability; Induction time; Coconut oil
Silpakorn U Science & Tech J
2 (2): 43-49, 2008
Introduction
Coconut oil has a long history for providing
human with several applications in daily life. It has
found use in foods (e.g., cooking oil), cosmetics (e.g.,
spa oil) and industries (e.g., lubricant). Coconut oil is
classified to unique group of vegetable oils called
lauric oils since its main composition is lauric
acid (Jayadas and Nair, 2006). Although most
compositions of coconut oil are saturated fatty acids,
its minor unsaturated fatty acids (oleic and linoleic
acids) can lead to oil rancidity because of lipid
oxidation. Generally, lipid oxidation is described by a
free-radical chain mechanism involving three steps,
i.e., (i) initiation or formation of initial free radicals
(RH J R• + H•), (ii) propagation of the free radicals
and formation of primary oxidation products such as
hydroperoxide (R• + O2 J ROO•, ROO• + RH J
ROOH + R•) and (iii) termination and formation of
secondary oxidation products such as carbonyl
compounds (R• + R• J R-R, R• + ROO• J ROOR,
ROO• + ROO• JO2 + ROOR) (Kanner and
Rosenthal, 1992; Gonzaga and Pasquini, 2006).
Oxidative stability is represented by the time in which
an oil sample resists to oxidation. This time, called
induction time or induction period, can be determined
by heating the oil sample to a constant temperature
Model of Coconut Oil DegradationSilpakorn U Science & Tech J Vol.2(2), 2008
44
and then measuring a parameter of the degree of
oxidation with time. Several physical and chemical
parameters can be used to elucidate the degree of
oxidation of the oil sample. The parameters related to
degree of oxidation remains practically constant
during the stability period of the oil. Afterwards, it
begins changing resulting in determination of the
induction time (Gonzaga and Pasquini, 2006). Many
experimental techniques have been applied to
investigate degree of oxidation of an oil sample, e.g.,
thermo-analytical method, electron spin resonance
spectroscopy, near infrared emission spectroscopy,
Fourier transform Raman spectroscopy, conductivity
measurement, fast ultrasound-assisted method and
UV absorbance measurement (Kanner and Rosenthal,
1992; Halbaut et al., 1997; Vieira and Regitano-d
´
Arce,
1998; Rudnik et al., 2001; Guillén and Cabo, 2002;
Cañizares-Macías et al., 2004; Gomez-Alonso et al.,
2004; Oliveira and Regitano-d
´
Arce, 2004; Velasco
et al., 2004; Muika et al., 2005; Gonzaga and Pasquini,
2006; Jayadas and Nair, 2006). Among these
techniques, UV absorbance measurement for
performing amounts of the secondary oxidation
products with time seemed to be a convenient method
in order to elucidate the induction time of an oil sample,
coconut oil. In addition, the obtained data could be
further studied for the model of degradation kinetics.
The aim of this work was to investigate the model
of degradation kinetics for coconut oil at various
heating temperatures.
Materials and Methods
Materials
Coconut oil was prepared in house by cold
process without refining, bleaching and deodorization.
This process is a common way of producing
commercial virgin coconut oil. Briefly, the gratings of
fresh and mature coconut were prepared using a
rotating grater. The wet gratings were put in a net
and extracted for coconut milk by squeezing. The
obtained milk was mixed with the coconut water and
allowed to settle through a process called culturing.
Afterwards, a three-layer mixture was obtained.
Protein contents rose up to the surface while oil and
water contents were in the middle and the lowest
layers, respectively. The oil was collected and filtered.
The gained oil was clear liquid with mild coconut
flavor. Chloroform was purchased from Lab-Scan
Analytical Science (Bangkok, Thailand) and used as
received without further purification.
Heating Coconut Oil at Various Temperatures
The 40 ml of coconut oil was heated at the
various temperatures of 130, 140, 145, 150 and
200 °C using a hot plate with a thermocouple for
controlling the temperature in an ambient condition.
The heated oil was taken and measured by UV
absorbance spectroscopy for its secondary oxidation
products occurring during 10 studying hours until the
color change from colorless to brown was observed
(Vieira and Regitano-d
´
Arce, 1998; Oliveira and
Regitano-d
´
Arce, 2004). The initial time (minute 0)
started when the oil reached to the determined
temperature. At each sampling time, 300 μl heated
coconut oil was taken. After the oil sample was
allowed to cool down to room temperature, it was
mixed with 3 ml chloroform. The heating experiment
was done in duplicate for each heating temperature.
Afterwards, the absorbance of each sample was
measured in duplicate using UV spectrophotometer
(Spectronic Genesys 5, Milton Roy, Ivyland, PA, USA)
at the wavelength of 270 nm while chloroform was
used as a blank. The studied wavelength was 270 nm
since changing in absorbance from the spectrum of
UV-scans between 200 to 400 nm of heated coconut
oil in the preliminary study was observed at 270 nm
(data not shown). This result was in accordance with
the report of Vieira and Regitano-d
´
Arce (1998) who
found that molecular friction during microwave
heating of refined canola, corn and soybean oils could
promote the formation of secondary oxidation
products (trienes and unsaturated ketones or
aldehydes) exhibiting absorption at 270 nm.
Silpakorn U Science & Tech J Vol.2(2), 2008W. Wongpoowarak et al.
45
Estimation of the Empirical Model of
Degradation Kinetics
The obtained data were analyzed to estimate the
empirical model in order to explain the degradation
kinetics of coconut oil. Due to the complexity of the
differential equation model and the data structures,
the non-linear regression of the differential equation
was manually performed in Microsoft ExcelTM.
Although this method was inferior to using standard
software since it did not provide confidence interval,
it was suffice to elucidate the possible kinetic model.
The finite differences numerical method was used
for numerically solving differential equation. In this
study, Euler method (order 1) was selected (Wylie
and Barrett, 1982). The time step used in the Euler
method was arbitrarily assigned and adjusted until
stable and converged prediction result was provided.
Afterwards, it was implemented as a user-defined
function in visual basic code for Microsoft ExcelTM
to calculate the incremental change described by the
differential equation for any time interval t1 J t2. The
criterion for performing non-linear regression was
to minimize Sum of Square Error (SSE) of the
prediction. The parameters of kinetics (degradation
rate constants) for each temperature were manually
adjusted until measurement data from all temperatures
could be simultaneously described by the differential
equations at each temperature, sharing only the initial
condition parameters (i.e., initial concentrations).
Results and Discussion
As shown in Figure 1, the absorbance values of
coconut oil heated at 130, 140, 145, 150 and 200 °C
were low and almost constant during the induction
time (i.e., 420, 210, 240, 120 and 15 min, respectively).
After the induction time, the absorbance values
abruptly increased due to the formation of secondary
oxidation products (Vieira and Regitano-d
´
Arce,
1998). In addition, higher slope of absorbance-
heating time curves during oxidation period at higher
heating temperature was observed in Figure 1. This
suggested that the heating temperature influenced not
only the oil stability, a decrease of the induction time,
but also the oxidation rate after the induction time
(Gonzaga and Pasquini, 2006).
Figure 1 Absorbance at 270 nm of coconut oil samples heated at various temperatures between 130 and
200 °C as a function of time.
0
0.5
1
1.5
2
2.5
3
3.5
4
0 100 200 300 400 500 600 700
Tim e ( mi n)
Absorbance
130 °C
140 °C 145 °C
150 °C
200 °C
Model of Coconut Oil DegradationSilpakorn U Science & Tech J Vol.2(2), 2008
46
The model was empirically constructed from the
results in Figure 1 by data fitting with various
scenarios and each scenario had its own specific sets
of differential equations to explain the degradation
kinetics of coconut oil as shown in Figure 2. Due to
the biphasic behavior in logarithmic scale, the
degradation of coconut oil could be described by two
simplified pathways, i.e., changing from non-degraded
coconut oil (A) to first (B) and second (C)
degradation products, respectively. The first pathway
(A J B) was called autocatalysis reaction since the
degradation product (B) could catalyze its own rate
of formation (i.e., positive feedback for its own
destruction). The autocatalysis model suggested that
oxidation possibly involved in degradation pathway
of coconut oil. The second pathway (A J C) was
simple first order reaction. These degradation
pathways could thus be described with the following
equations (Martin, 1993):
Figure 2 Model of degradation kinetics of coconut oil. A refers to non-degraded coconut oil. B and C refer to
degradation products of coconut oil. The k1 and k2 refer to rate constants.
d[A]/dt = -k1[A][B]-k2[A] (1)
d[B]/dt = k1[A][B] (2)
d[C]/dt = k2[A] (3)
where [A] was concentration of non-degraded
coconut oil; [B] and [C] were concentrations of
degradation products of coconut oil; k1 and k2 were
rate constants of autocatalysis and first order
reactions, respectively. The concentration could be
expressed as UV absorbance; therefore absorbance
values were used instead of concentration throughout
this experiment.
The obtained data from the experiment were
compared with the calculated data as presented in
Figure 3. The measured and predicted values were in
the same tendency while the heating temperatures
were lower than or equal to 150 °C. Hence, the
proposed model could properly explain the
degradation of coconut oil in the range of low
AB
C
k1
k2
k1
k2
AB
C
Silpakorn U Science & Tech J Vol.2(2), 2008W. Wongpoowarak et al.
47
Figure 3 Plot of relationship between log (absorbance) of heated coconut oil samples (y axis) at various
temperatures and time (x axis). The dotted and hard lines represent the measured and predicted
values, respectively. The light lines show the partial effect of A, B and C estimated from the nonlinear
regression of the model.
temperature (≤150 °C). However, the model could
partially explain the data at higher heating
temperature (200 °C) in first 60 minutes since there
might be further degradation that contributed to
absorbance far exceed that found in lower
temperature range. This result suggested that
degradation of coconut oil at high heating
temperature (200 °C) may occur in different pathways
from that at low heating temperature (≤150 °C).
The best-fit values of k1 and k2 could be found
by manual optimization in Microsoft ExcelTM
Worksheet. All the regression parameters were
manually optimized to achieve the best fit criterion
(i.e., minimize SSE) for each and all particular
temperatures. The initial values of all components
(i.e., Ao, Bo, Co) were fixed as the same values for all
temperature data and varied particular sets of k1 and
k2 for each temperature.
Model of Coconut Oil DegradationSilpakorn U Science & Tech J Vol.2(2), 2008
48
The induction time could be roughly estimated via
graphical method. However, in order to standardize
the data interpretation process, the degradation
profile was simulated to predict induction time.
Before the induction time, the slope was low
(non-zero slope) and could be approximated by a
constant slope. After the induction time, the slope was
at its highest and could be approximated by another
constant slope. From this observation, the derivative
of the slope (i.e., the second derivative of the
concentration with time) should be highest at the
induction time. We simulated the system and
computed second derivative of the degradation curve
in respect to the heating time; the time to reach
the highest second derivative of the simulated
concentration of the degradation products was the
induction time in this study.
heating temperature (200 °C) due to different
degradation behavior. It could be noted in Table 1
that higher values of percentage of prediction error
for induction time compared to the experimental
data in Figure 1 were observed at higher heating
temperatures. Theoretically, the k1 and k2 could be
further used to predict the shelf-life of the sample using
Arrhenius equation (Martin, 1993). However, the
actual kinetic processes were suspected to be more
complex than described by UV measurement in this
study. Hence, only induction time extrapolation was
used in this study.
Conclusions
The obtained empirical model suggested that there
were at least two types of degradation pathways
involved while coconut oil was heated. The first-type
Table 1 The rate constants (k1 and k2), predicted induction time and percentage of prediction error of coconut
oil degradation at various heating temperatures.
Heating k1 k2 Predicted Induction % Prediction
Temperature (°C) (absorbance-1 min-1) (min-1) Time (min) Error
130 0.019 7.9 x 10-6 467 11.19
140 0.039 8.1 x 10-5 234 11.43
145 0.030 1.7 x 10-4 282 17.50
150 0.060 6.7 x 10-4 157 30.83
200 0.300 7.0 x 10-3 37 146.67
The summary of predicted k1, k2 and induction
time values at various heating temperatures are
shown in Table 1. From the obtained model, it could
be estimated from linear regression between
temperature and logarithm of induction time that
induction time decreased approximately by half for
every increase of 14 °C in the heating temperature. In
previous study, Gonzaga and Pasquini (2006) found
that the induction time of canola oil decreased
approximately by half for every increase of 10 °C in
the heating temperature. This model was not suitable
for describing the degradation of coconut oil at high
of pathway was autocatalysis (second order kinetics)
due to positive feedback from the primary oxidation
products. The second-type of pathway was first
order kinetics. Each pathway might consist of various
smaller kinetic processes of the same nature. From
simulation study, autocatalysis was found to be very
slow in the beginning and very fast after reaching
critical level at induction period. Since autocatalysis
was positive feedback process (i.e. self-destruct
feedback), preventing its initiation should also prevent
its destruction process.
Silpakorn U Science & Tech J Vol.2(2), 2008W. Wongpoowarak et al.
49
References
Ca
ñ
izares-Macías, M. P., García-Mesa, M. D., and
Castro, M. D. (2004) Fast ultrasound-assisted
method for the determination of the oxidative
stability of virgin olive oil. Analytica Chimica
Acta 502: 161-166.
Gomez-Alonso, S., Mancebo-Campos, V., Salvador,
D., and Fregapane, G. (2004) Oxidation
kinetics in olive oil triacylglycerols under
accelerated shelf-life testing (25-75 °C).
European Journal of Lipid Science and
Technology 106: 369-375.
Gonzaga, F. B. and Pasquini, C. (2006) A new method
for determination of the oxidative stability of
edible oils at frying temperatures using near
infrared emission spectroscopy. Analytica
Chimica Acta 70: 129-135.
Guill
é
n, M. D. and Cabo, N. (2002) Fourier transform
infrared spectra data versus peroxide and
anisidine values to determine oxidative stability
of edible oils. Food Chemistry 77: 503-510.
Halbaut, L., Barb
é
, C., Ar
ó
ztegui, M., and Torre, C.
(1997) Oxidative stability of semi-solid
excipient mixtures with corn oil and its
implication in the degradation of vitamin A.
International Journal of Pharmaceutics
147: 31-40.
Jayadas, N. H. and Nair, K. P. (2006) Coconut oil as
base oil for industrial lubricants - evaluation
and modification of thermal, oxidative and
low temperature properties. Tribology
International 39: 873-878.
Kanner, J. and Rosenthal, I. (1992) An assessment of
lipid oxidation in foods. Pure and Applied
Chemistry 64(12): 1959-1964.
Martin, A. (1993) Physical Pharmacy, Lea &
Febiger, Philadelphia.
Muika, B., Lendl, B., Molina-Diaz, A., and Ayora-
Cañada, M.J. (2005) Direct monitoring of lipid
oxidation in edible oils by Fourier transform
Raman spectroscopy. Chemistry and Physics
of Lipids 134: 173-182.
Oliveira, J. T. G. and Regitano-d
´
Arce, M. A. B. (2004)
Determining economical TBHQ doses for corn
oil stability. Ci
ê
ncia e Tecnologia de Alimentos
24(3): 413-418.
Rudnik, E., Szczucinska, A., Gwardiak, H., Szulc, A.,
and Winiarska, A. (2001) Comparative studies
of oxidative stability of linseed oil.
Thermochimica Acta 370: 135-140.
Velasco, J., Andersen, M. L., and Skibsted, L. H.
(2004) Evaluation of oxidative stability of
vegetable oils by monitoring the tendency to
radical formation: a comparison of electron spin
resonance spectroscopy with the Rancimat
method and differential scanning calorimetry.
Food Chemistry 85: 623-632.
Vieira, T. M. F. S. and Regitano-d
´
Arce, M. A. B.
(1998) Stability of oils heated by microwave:
UV - spectrophotometric evaluation. Ci
ê
ncia
e Tecnologia de Alimentos 18(4): 433-437.
Wylie, C. R. and Barrett, L. C. (1982) Advanced
Engineering Mathematics, McGraw-Hill,
Singapore.
