Palm oil is a flesh fruit fat with a distinctive orange-red colour, due to its high content of carotenoids. Palm oil being a major source of vegetable fats is used extensively in cooking, cosmetics and in the manufacture of soap. Crude palm oil is found to contain pigment likes carotenes and impurities like free fatty acids. When processing vegetable oils, the bleaching operation is an important step intended to remove pigments and other unwanted constituents such as those of mucilaginous character and other volatiles (Christidis et al., 1997). Many types of adsorbent have been tested for the removal of pigments from vegetable oil (Kamga et al., 2000; Falaras et al., 2000; Topallar, 1998; Proctor and Snyder, 1988; Kheok and Lim, 1982). Adsorption materials used nowadays throughout the world by vegetable oil refiners are mainly activated montmorillonite clays. Christidis and Kosiari (2003) showed that removal of β-Carotene from crude maize oil with acid activated low grade bentonite from Cyprus is a chemical adsorption process. The isotherm they obtained was of the Freundlich type. None of the above studies mentioned the adsorption of free fatty acids during the decolourisation process. Moreover other authors pointed that the bleaching process led to and increased of the free fatty acids content of the vegetable oil (Habile et al, 1992; Boki et al., 1992; Nnandozie et al., 1989). More recently (Bike Mbah et al., 2005), we show that acid activated clay from Cameroon could be used successfully for the adsorption of pigments and free fatty acids of shea butter. The high value of the activation energies obtained indicated that the adsorption of shea butter pigment was of a chemical nature. It is the purpose of this contribution to study the mechanism of the simultaneous adsorption of carotene and free fatty acids of palm oil, by means of kinetics, with determination of activation energy and construction of the adsorption isotherm.
MATERIALS AND METHODS
Palm oil used for this research was the olein portion of fractionated palm oil provided by SOCAPALM. Co (Cameroon). The physicochemical characteristic of the crude fat was: Acid index 12.34, Peroxyde value 96 meq kg1, Absorbance at 445 nm 1.975.
Adsorbents used were activated clays and an industrial adsorbent. The industrial
adsorbent was obtained from Engelhard Co (Netherland). The clays were extracted
from aggregates of soil collected at Kaélé which is a locality
in the far north province of Cameroon (latitudes 10° and 10°15' North
and longitudes 14°10' and 14°35' East).
|LD: Lower than the detection limit; ND: Not Determined
The mineral composition of the natural clays was: smectites 74.4, kaolinite
7.7, quartz 15.4 (Nguetnkam et al., 2005). Other adsorbent characteristic
are presented in Table 1.
The extraction of the argillaceous fraction was made by gravimetric sedimentation. A sample of soil (200 g) was placed in a beaker and to this was added 200 mL of distilled water. The mixture was then placed in an ultrasound water bath for 25 min and transferred thereafter into a sedimentation vessel of 1500 mL volume. Distilled water was added up to 1200 mL. The mixture was then stirred for few minutes and left to stand for 8 h. After this time, the mixture located a higher of 10 cm from edge of the sedimentation vessel was collected by siphon. According to Stokes law this mixture should contain the argillaceous fraction whose particles size is lower than 2 μm. The process of agitation-rest-siphon was repeated several times. The mixture of clay obtained was placed in an oven at 105°C until completely evaporation of water.
The procedure of activation was carried out according to the method described by Vincente-Rodriguez et al. (1996). The clay sample (50 g) was introduced into a beaker and 250 mL of sulphuric acid (analytical grade) solution added; three different concentrations of sulphuric acid solution were used, that is 0.5, 1.0 and 2.0 M. The mixture was homogenized from time to time at ambient temperature using a glass rod for 3 h. At the end of this period, the mixture was washed several times with distilled water until the silver nitrate test for sulphate ion was negative.
Nitrogen adsorption-desorption isotherms at 77 K were recorded on a step-by-step automatic home-built setup (Laboratoire Environnement et Minéralurgie; Nancy, France). Specific Surfaces Areas (SSA) were determined from adsorption data by applying the Brunauer-Emmet-Teller (BET) equation. The error in the determination of SSA was estimated as±1 m2 g1. Pore size distributions were calculated on the desorption branch using the numerical integration method of Barrett-Joyner-Halenda, assuming slit-shaped pores.
For the adsorption experiment, 5 g of palm oil was placed in a stoppered conical flask held in a shaking (80 cycles min1) water bath thermostated at a predetermined temperature, that is 65±1, 80±1, or 90±1°C. When the content had reached the required temperature, usually not more than 15 min, a known amount of adsorbent was added to the flask, which was then shaken for the desired time. For the kinetic studies residence time varied between 5 and 75 min. Thereafter, the contents of the flask were filtered through Whatman No. 1 filter paper and the concentration of carotene and free fatty acids in the filtrate determined as described below. The amount of pigment adsorbed was determined by difference between the initial and the final concentration, while the amount of free fatty acids adsorbed was determined by difference between the initial and the final acid value. The amount of adsorbent added varied between 1 and 5% (w/w) with respect to the palm oil. In all cases the adsorbent was previously dried overnight at 105°C. Each adsorption experiment was repeated twice, the results reported are de means of three measurements. The error in the determination of amount of pigment and fatty acids adsorbed was evaluated to 10%.
The evaluation of the amount of pigment removed was made by UV-visible spectroscopy. The sample 0.1 g of palm oil was diluted in 7.5 mL of petroleum ether (analytical grade) and the absorbance of the sample determined at 445 nm using petroleum ether as reference.
The acid value and peroxide value was determined according to the respective AFNOR method (1981).
RESULTS AND DISCUSSION
Determination of clay yield from soil aggregate: The percentage yield of clay from the soil aggregate was 40.3±0.5%. This value lies within the range suggested by Nnandozie et al. (1989).
UV-visible spectrum of palm oil: The adsorption spectrum of crude palm oil (result not shown) presents a maximum adsorption band at 445 nm, with shoulders at 420 and 470 nm. The reduction of the optical density to this maximum of absorption was used to follow the discoloration process.
The thermal stability of the palm oil pigment was tested. For this purpose,
crude fat was heated at 65, 80 and 90°C for 60 min and the absorbance of
the final product read at 445 nm.
||Kinetics of the adsorption of palm oil carotene at various
temperatures, 90°C, 80°C and 65°C, by
EN (a), A1M (b), A0.5M (c) and A2M (d) clays
It appears that there is no major change in the pigment concentration since
the absorbance remains almost constant through out the heating process.
Kinetics of the adsorption of palm oil carotenes: Figure
1 shows the kinetics of the adsorption of palm oil carotenes at three different
temperatures (90, 80, 65°C) by industrial clay (EN, Fig. 1a),
the local clay activated with sulphuric acid solution at concentrations of 1M
(A1 M, Fig. 1b), 0.5M (A0.5 M, Fig. 1c)
and 2M (A2 M, Fig. 1d). A clay dosage of 3% w/w oil was used
for the kinetic studies. It is evident from these results that the time
required to reach the adsorption equilibrium decreases as the temperature increases,
regardless of the adsorbent used. It was observed indeed that at 65°C the
time required to reach the adsorption equilibrium is 80 min for all the adsorbents
used. At 80°C this time reduce to 30 min for EN clay, 45 min for A0.5M and
A1M clays and finally 60 min for A2M clay. At 90°C the time require for
the adsorption equilibrium is 20 min, 30 min and 50 min for EN, A0.5M and A1M
and A2M clays, respectively. This observed reduction correlates well
with the enhancement of oil viscosity with decreasing temperatures reported
by Langmaack and Eggers (2002). When the viscosity of the oil is high, there
will be a slowing down of transport mechanisms due to decreasing diffusion rates
of pigment molecules to the surface of the adsorbent. Other authors attributed
the increase in adsorption rate with temperature to the activation of more adsorption
sites by heat (Achife and Ibenesi, 1989).
It was also observed that the amount of pigments adsorbed increased with the
temperature independent of the adsorbent used. However the clays had different
adsorbing capacities. The difference in the bleaching efficiency of local clays
can be attributed to their structure. Preliminary tests showed that natural
clays were ineffective in the adsorption of palm oil pigment. A0.5M clay had
a better performance compared to natural clay; this can be explained by the
increase in surface area and pores volumes during acid leaching (Yebra-Rodrigues
et al., 2003; Nnandozie et al., 1989; Achife and Ibenesi, 1989).
In effect A0.5M had a greater surface area (143.6 m2 g1)
than natural clay (110.9 m2 g1). The efficiency
of A0.5M clay is lower than that of A1M clay; this is due to the fact that during
acid leaching with 0.5 M sulphuric acid not all of the adsorption sites were
liberated because of the weak concentration of the acid solution. Although A2M
clay have greater surface area (202.0 m2 g1) than
A1M (168.0 m2 g1) and A0.5M (143.6 m2
g1) clays it had the lowest bleaching efficiency. This poor
performance of A2M clay can be attributed to the collapse of the crystalline
structure of the clay with the formation of silicium and aluminium oxide during
the severe acid leaching (Nguetnkam et al., 2005; Christidis et al.,
1997; Kheok and Lim, 1982). Preliminary studies with silicium and aluminium
oxide showed that there was an insignificant amount of pigment adsorbed from
palm oil. The evolution of the bleaching efficiency of clays with the increases
of acid concentration of the acid solution used for clay leaching is similar
to those obtained by Biké Mbah et al. (2005).
Estimation of activation energy: According to Brimberg (1982) the speed
of discoloration of vegetable oils is given by the equation ln(C/Co) = -k(t)0.5
where t is the time of contact between the adsorbent and oil, C the concentration
of the pigments at time t, Co the initial concentration of the pigments and
k the rate constant. According to the Beer-Lambert law the absorbance is proportional
to the concentration of the pigments in the oil, therefore the equation of Brimberg
can take the form ln(A/Ao) = -k(t)0.5 where A is the absorbance of
the oil bleached at time t and Ao the absorbance of crude oil. From this equation
the linear regression between ln(A) and (t)0.5 is a straight line
whose slope is equal to k. The values of the rate constant k for palm oil decolourised
with various clays at three different temperatures are given in Table
||Rate constants and activation energies for the adsorption
of palm oil carotenes on various adsorbents at different temperatures
It is can be deduced from this table that the rate constants vary from 0.018
min1 for A2M to 0.339 min1 for EN. For all
the clays used, the smallest rate constants were obtained at 65°C. This
is in conformity with the results obtained in the kinetics of discoloration
according to which, at low temperature, the adsorption of the pigments of the
palm oil is very slow.
The activation energy can be deduced from these values of k using the Arrhenius
equation k = k0exp(-Ea/RT) where k is the rate constant, k0
the Arrhenius constant, Ea the activation energy, R the perfect gas constant
and T the absolute temperature. From this equation the linear regression
between ln(k) and 1/T is a straight line whose slope is equal to -Ea/R. The
values of the activation energy obtained for each adsorbent used are also reported
in Table 2. These values indicate that the adsorption of palm
oil carotenes on the various clays used is probably a chemical adsorption process.
This result is in agreement with that found by Christidis and Kosiari (2003)
for the adsorption of β-carotene from maize oil.
Figure 2 shows the variation in the amount of free fatty
acid adsorbed with time at 90°C. Adsorption equilibrium is attained at about
40 min for EN clay and at about 60 min for the activated clays. It can be seen
that the industrial adsorbent EN clay, is the most effective adsorbent for free
fatty acid, followed by A1M, A0.5M and A2M activated clays, respectively. The
range of acid concentrations used to treat the clays therefore has a significant
influence on their adsorption capacities with respect to the adsorption of free
fatty acids. The results presented here on the adsorption of free fatty acids
are not in agreement with the results reported in previous works. Indeed, it
is generally found or admitted in the literature that the acid value of vegetable
oils increases during discoloration (Boki et al., 1992; Habile et
al., 1992; Nnandozie et al., 1989). This difference could be related
to either the nature of the adsorbents, or to the temperatures at which the
experiment was performed. The works cited in the literature were generally completed
at temperatures higher than those implemented in this research.
Evolution of peroxide value: The peroxide value for palm oil was determined in the course of discoloration at 90°C.
Kinetics of free fatty acid
adsorption by the various adsorbent
A1 M activated clay, (
A0.5 M activated clay and (
||Evolution of peroxide value with time. EN,
A0.5M and A2M,
Fig. 3 shows that during discoloration, the products of oxidation
are also eliminated. It appears that the adsorbents which are effective in the
adsorption of the pigments and free fatty acids are also those which are effective
in the destruction of peroxide. The order of effectiveness is the same. This
result is in agreement with that of Kheok and Lim (1982), but in disagreement
with the results obtained by other authors (Nnandozie et al., 1989).
Adsorption isotherms of carotene: The kinetic study has shown that the
required temperature and time to reach adsorption equilibrium are 90°C and
30 min, respectively. These parameters were selected to build the adsorption
isotherms. The relative amount of pigment adsorbed per gram of adsorbent X and
the concentration of pigment at equilibrium Xe are obtained by the following
equations: X = ((Ao-A)/(Ao))/m and Xe = A/Ao where Ao and A are the optical
densities of the crude and decolourised palm oil, respectively and m is the
amount of adsorbent used.
||Parameters of Freundlich equation for the adsorption of palm
||Adsorption isotherms of carotene on the various adsorbents.
A0.5M and ()
The adsorption isotherms obtained are presented in Fig. 4.
It is observed that the amount of pigments adsorbed per gram of adsorbent increases
according to the amount of clay used.
Modeling of the isotherms: The models frequently used to represent the adsorption isotherms of pigments from vegetable oils or from solvents are the Langmuir and Freundlich equations (Christidis and Kosiari, 2003; Kamga et al., 2000; Boki et al., 1992; Achife and Ibenesi, 1989; Sarier and Güler, 1988). Thus, we tested these two equations with our experimental data. The correlation coefficients for Langmuir equation were 0.004, 0.626, 0.664 and 0.900 for EN, A0.5M, A1M, A2M clays, respectively. It was therefore deduced that the Langmuir equation does not fit our experiment data.
The equation of adsorption of Freundlich can be given by the equation:
log(X/m) = log(λ) + (1/n)log(Xe)
where, λ is the constant reflecting the measurement of the capacity of
adsorption and n the constant reflecting the affinity of the pigments for the
adsorbent. X/m is the relative amount of pigments adsorbed per gram of adsorbent
and Xe the relative amount of pigment at equilibrium. The values of the correlation
coefficients obtained (Table 3) show that the Freundlich equation
is applicable to the adsorption of palm oil carotenes by clays.
Adsorption of carotene and free fatty acids from crude palm oil by local clay, activated with various concentrations of sulphuric acid (0.5, 1 and 2 M) was investigated. The kinetic study of discolouration showed that the adsorption of the pigments increased with temperature. The most effective adsorbents for discolouration were clays EN and A1M followed by clays A0.5M and A2M. An optimal effectiveness of local clay seemed to be obtained when it was activated with solution of sulphuric acid of 1 M concentration. The most efficient adsorbents for the adsorption of the pigments were also those which gave the best results for the adsorption of free fatty acids and in the elimination of peroxides.
After physicochemical analysis, it appeared clearly that discolouration not only makes it possible to give an attractive colour to the palm oil, but also to have stable oil and of good quality. Indeed, free fatty acids are adsorbed on clays and peroxides contained in raw oil are destroyed during the process.
The high values of the energies of activation determined according to the model of Brimberg suggest that the adsorption of palm oil pigments on various clays is probably a chemical process. The adsorption isotherms obtained follow Freundlich equation for the adsorption of pigments from palm oil.
The improvement of the quality of palm oil obtained after discolouration, show that activated local clay can be used for palm oil bleaching.
show that acid activated clay from Cameroon could be used successfully for the adsorption of pigments and free fatty acids.
The authors are grateful to Dr. F. Villiéras (LEM-INPL Nancy, France) for the surface area determination of the clays sample and for fruitful discussions.