Journal of Colloid and Interface Science 235, 93100 (2001)

doi:10.1006/jcis.2000.7175, available online at http://www.idealibrary.com on

KaolinCarbon Adsorbents for Carotene Removal of Red Palm Oil

Mohd Zobir Bin Hussein,1 Dzulkefly Kuang, Zulkarnain Zainal, and Tan Kian Teck
Multifunctional Materials for Industrial Application (MULIA) Research Group, Department of Chemistry, Universiti Putra Malaysia,
43400UPM, Serdang, Selangor, Malaysia
Received April 26, 2000; accepted August 21, 2000

Kaolincarbon adsorbents were prepared with and without sulfuric acid pretreatment followed by activationcarbonization at
500 C. The total surface area of the resulting kaolincarbon adsorbents was found to be decreased with the increase in kaolin
loading. Sulfuric acid pretreatment of the precursor prior to the
carbonizationactivation processes resulted in the enhancement of
total surface area but reduced the micropore surface area of the
resulting adsorbents. At the same time, this improved carotene adsorption capacity from red palm oil. However, recovery of carotene
from the carotene-adsorbed adsorbent is only improved when the
acid pretreatment of the precursor was done at a high loading percentage of activated carbon. Similarly, the peroxide value (PV)
increased. A maximum removal of carotene from red palm oil
was obtained at 20% kaolin loading for both adsorbents prepared
with and without sulfuric acid pretreatment with about 45 and
65% carotene removal, respectively, from a 30-ppm solution. This
indicates that pretreatment with sulfuric acid, prior to the
activationcarbonization process, increased the carotene uptake by
the resulting adsorbent. However, a further increase in the kaolin
loading resulted in the decrease of carotene removal. About 34%
of carotene adsorbed can be recovered from both types of adsorbents under optimum condition, in which the percentage recovered
decreased with the increase in kaolin loading. On the other hand,
the PV increased with kaolin loading at around 5464 mEq/kg for
both types of adsorbents. It was also found that carotene uptake
by the adsorbents is high if the adsorbent contains a high percentage of activated carbon. Similarly, carotene recovery is high
and less oxidation can be observed, as indicated by the lower PV
value. C 2001 Academic Press
Key Words: kaolin; activated carbon; activated carbonkaolin
complexes; surface area and porosity; carotene removal; red palm
oil.

Activated clay, commonly termed fullers earth, is the most

popular adsorbent for decoloration of edible oil because it is
cheaper compared to activated carbon. As Malaysia is currently
one of the worlds major producers of palm oil, expansion of this
industry was followed by the generation of an enormous amount
of by-products, especially at plantation grounds, oil mills, and
refineries. Given the current scale of industrial waste generation
and stringent control on industry to minimize the environmental
impact, research into the production of adsorbents from this type
of industrial wastes has considerable potential.
Another cheap and easily available raw material is clay. Its
modification for adsorbent purposes can be done using mineral
acid treatment and the resulting activated clay is useful for color
removal. The combination of the activated clay with activated
carbon is hoped to enhance the property of the resulting materials for separation purposes. Previous studies show that the
carbonization of spent bleaching clay from the vegetable oil
processing industry has promising results for use as adsorbents
(5, 6). In addition, claycarbon adsorbent from fullers earth
and alkaline aqueous tarry industrial waste was also prepared
and showed good adsorption on phenol compounds (7).
Here we discuss our preliminary work on the preparation of
a mixture of kaolin-activated carbon (KAM) and its complexes
(KAC). Kaolin and activated carbon, purchased from Sigma and
from a local manufacturer, respectively, were used as a source
of raw materials. Sulfuric acid pretreatment of the precursor before the carbonizationactivation process was done to prepare
KACs. This treatment is hoped to improve the resulting physicochemical properties of the adsorbent, especially for the purpose
of color removal, in particular -carotene from red palm oil
(RPO).

INTRODUCTION

MATERIALS AND METHODS

Activated clay and activated carbon are two common adsorbents, which have many applications in separation technology,
ranging from the food industry to waste water treatment. Of
late, the preparation of both has been a subject of interest, especially from low-cost agroindustrial by-products such as palm
kernel shells (1), oil palm trunks (2), almond shells (3), and olive
stones (4), to name a few.
1

Preparation of Kaolin-Activated Carbon Mixture

Kaolin-activated carbon mixture was prepared by adding
10 g of activated carbon of coconut shell origin with 20%
(w/w) kaolin. Before the mixing process, the activated carbon was ground and sieved with an Endicott 150-m sieve
and oven-dried overnight. The resulting mixture of kaolin
carbon was moistened with distilled water and compressed into
pallets.

To whom correspondence should be addressed.

93

0021-9797/01 $35.00

C 2001 by Academic Press

Copyright
All rights of reproduction in any form reserved.

94

HUSSEIN ET AL.

The pallets were then placed inside the center of a quartz

column, which sits vertically inside a tubular electric furnace.
The activation temperature was set at 500 C under an inert flow
of nitrogen gas for 3 h. The resulting product was left to cool
to room temperature and kept in a sample bottle for further
treatment or characterization. The same procedure was repeated
for various percentages of kaolin loading, 0100%.

method (9) and was calculated by the formula

25 383/[100W (as ab )],

[1]

where W is the weight of the sample in grams and as and ab are

the absorbance of the sample and cuvette error, respectively.
Desorption of Carotene from the Adsorbents

Preparation of Activated KaolinCarbon Complexes

KACs of various percentages of kaolin loading were prepared
as follows. Ten grams of powdered activated carbon of coconut
shell origin was prepared as was described earlier and was added
into a conical flask. About 20% (w/w) of kaolin and 100 ml of
2 M sulfuric acid was added. The resulting slurry was aged
at 70 C overnight or until dryness, followed by oven drying
overnight.
The sample was placed at the center of a quartz column and
inside a tubular electric furnace. The activation temperature was
set at 500 C under an inert flow of nitrogen gas for 3 h. The
resulting product was left to cool to room temperature and kept
in a sample bottle for further treatment or characterization. Other
samples of various kaolin loading were also prepared in a similar
manner.
Physicochemical Characterizations of the Adsorbents
BET surface area, pore volume, and nitrogen adsorption
desorption isotherms of the samples were obtained using a Micromeritics ASAP 2000 surface area analyzer at the temperature
of liquid nitrogen. Pore volume and pore size distribution were
determined using the Barret, Joyner, and Halenda (BJH) desorption method (8). Each carbon sample was degassed at 473 K
under vacuum at least overnight.
Color Removal of Red Palm Oil
Red palm oil purchased from a local market was used in this
study. The red-colored oil, usually used for cooking, is rich in
carotene and was used without further treatment.
About 2 g of RPO was dissolved in 40 ml of hexane in a
conical flask, and 0.2 g of the adsorbents KAM or KAC was
added into the mixture. The slurry was stirred for 40 min with
a magnetic stirrer at room temperature. The filtrate was collected and hexane was removed under vacuum with a rotary
evaporator to recover the RPO. The carotene left in the oil after
extraction was determined by diluting the RPO with hexane and
the absorbance at 446 nm was measured by a Shidmadzu-160A
spectrophotometer. The percentage of carotene adsorbed was
calculated by comparing the difference between the carotene
left in the solution to that in the initial RPO.
The carotene content of RPO was defined and calculated as carotene in parts per million (ppm). The carotene concentration
in the oil sample was determined according to the PORIM test

To recover the adsorbed carotene, the carotene-adsorbed adsorbents were mixed with 50 ml of toluene and stirred continuously for 40 min, followed by solvent filtration and evaporation
using a rotary evaporator. The recovered carotene obtained was
subjected to UV-visible measurement and the amount was calculated using Eq. [1].
Determination of Peroxide Value
The extent of oxidation of RPO after carotene removal was
determined by its peroxide value. About 0.5 g of RPO was added
into a 250-ml conical flask, followed by the addition of 30 ml
of a mixture of acetic acidchloroform solution (3:2 v/v). The
resulting solution was then added with 0.5 ml of KI saturated
solution and swirled for 1 min. After that, 30 ml of distilled
water was added, followed by titration with standardized sodium
thiosulfate solution using constant and vigorous shaking until the
light-yellow color has almost disappeared. A total of 0.5 ml of
1% starch solution was added as an indicator and the titration
was continued until the blue color disappeared. The result was
reported as milliequivalents of peroxide per kilogram of sample
(mEq/kg).
RESULTS AND DISCUSSION

X-Ray Diffraction
Figure 1a shows powder X-ray diffraction (PXRD) patterns
for activated carbon and its mixtures with kaolin at various
kaolin-loading percentages, 20100%. As shown in the figure,
an amorphous character of the activated carbon can be seen
clearly for 100% activated carbon. On the other hand, the XRD
pattern for 100% kaolin is dominated by two peaks with d values

of 7.14 and 3.55 A.

A mixture of the property of activated carbon and kaolin can
be seen in the XRD patterns of KAMs, at kaolin loading of
2080%, as shown in the figure. A mixture of an amorphous
and crystalline pattern can be seen clearly up to 40% of kaolin
loading. However, the XRD patterns of the amorphous phase
are slowly diminished, resulting in XRD patterns dominated
by the kaolinite phase, as shown by KAM with kaolin content
>60%.
Figure 1b shows the XRD patterns for KAC. In comparison
with the XRD patterns of KAM, it shows that the XRD pattern of KAC is dominated by the kaolinite phase, even at low a
loading percentage of kaolin, 20%. This shows that activation
and carbonization of kaolin-activated carbon mixture with 2 M

ADSORBENTS FOR CAROTENE REMOVAL OF RED PALM OIL

95

sulfuric acid somehow enhanced the crystallinity property of the

resulting KAC.
All of the XRD patterns for KAC are dominated by the peak
The peak at 7.14 A
is still present,
with d value of 3.55 A.
indicating that the basal spacing of the kaolinite remains the
same after treatment with sulfuric acid followed by heating at
500 C for 3 h. In addition, many more peaks with fairly high
intensity can be seen, compared to the XRD pattern for kaolin.
This is presumably due to the formation of more crystalline
phases of the resulting materials.
Surface Area and Porosity
In order to study the effect of surface properties of the resulting
materials upon carbonizationactivation and at various loading
percentages of kaolin, we measured the surface area and pore
size distribution using the nitrogen gas adsorptiondesorption
technique.
AdsorptionDesorption Isotherms
Figure 2 shows the adsorptiondesorption isotherms for KAM
at various kaolin loading, 0100%. As shown in Fig. 2a, the adsorption isotherm of kaolin is dominated by Type IV (8), indicating a mesopore-rich texture of the silicate layered material.
On the other hand, the adsorption isotherm for activated
carbon is Type I (Fig. 2f) by BDDT classification (8), with

FIG. 1. PXRD patterns for (a) KAM and (b) KAC prepared at various
percentages of kaolin loading (w/w), 0100%; the peak labeled with a diamond
is due to an aluminum sample holder.

FIG. 2. Adsorptiondesorption isotherms for KAM prepared at various percentages of kaolin loading (w/w): (a) 100, (b) 80, (c) 60, (d) 40, (e) 20, and
(f) 0%.

96

HUSSEIN ET AL.

ments. It is noteworthy also that there is a higher slope of the

adsorption isotherm at high relative pressure, >0.2, indicating
a higher percentage of mesopore in the activated carbon prepared by treating it with 2 M sulfuric acid followed by heating
compared to that without any acid treatment.
The adsorptiondesorption isotherms for KACs at various
kaolin loading is shown in (Figs. 3b3e). As shown in the figure,
the general aspect of the adsorption isotherm is dominated by
Type II (8), indicating the mesopore-rich texture of the adsorbent
materials. However, the presence of Type IV, which belongs to
kaolin, cannot be ruled out, which can be seen clearly, especially
at high relative pressure, near to 1. The domination of Type II in
the resulting KACs is due to the high contribution of the surface
property of activated carbon.
Surface Area

FIG. 3. Adsorptiondesorption isotherms for KAC prepared at various percentages of kaolin loading (w/w): (a) 100, (b) 80, (c) 60, (d) 40, (e) 20, and
(f) 0%.

adsorption increasing fairly rapidly at low relative pressure in

a range of 0.00.2, followed by a slow uptake of the adsorbent
at a relative pressure of >0.2. This indicates that the activated
carbon used in this study is essentially microporous. However, a
small portion of mesopore is also present, which can be deduced
from the presence of the slope of the adsorption isotherm at high
relative pressure, >0.2.
The adsorption isotherms for KAM prepared at various kaolin
loading (Figs. 2b2e) show a combination of both Type I and
Type IV isotherms. The character of the isotherms is dominated
by Type I if the loading percentage of activated carbon is high
and by Type IV if the loading percentage of kaolin is high.
Figure 3 shows the adsorptiondesorption isotherms for KAC,
which was prepared by activation of a kaolin-activated carbon
mixture with 2 M sulfuric acid, followed by heating at 500 C
for 3 h. As shown in Fig. 3a, the general aspect of adsorption
isotherm for kaolin still remains, but a Type IV with more open
hysterisis was obtained, indicating that the surface property of
the resulting material changed slightly.
For activated carbon, the adsorption isotherm changes
slightly, in which the Type I transforms to Type II with a wider
hysteresis loop compared to the precursor. This indicates the
formation of more mesopores as a result of acid and heat treat-

As a result of the adsorption of nitrogen gas on KAM and

KAC, total and micropore surface areas together with micropore volume and average pore diameter of the adsorbents were
obtained. These are given in Table 1.
The total surface area of activated carbon and kaolin is 1783
and 19 m2 g1 , with a micropore area of 940 and 2 m2 g1 ,
respectively. Upon treatment with 2 M sulfuric acid followed by
activation at 500 C for 3 h, the surface area increased to 1860 and
37 m2 g1 , but the micropore area decreased to 828 for activated
carbon and increased to 3 m2 g1 for kaolin. This indicates that,
for activated carbon, the treatment increased the total surface
area but the micropore surface area is actually decreased. On
the other hand, for kaolin both total and micropore surface area
increased with treatment.
On addition of kaolin at various kaolin-loading percentages
in KAM, the surface area decreased with the percentage almost
linearly, as shown in Fig. 4. A similar trend was also observed
for the micropore area. However, the percentage contribution of
the micropore area to the total surface area has a different trend,
in which it is almost constant up to 80% of kaolin loading,
after which it decreased to about 10%. This is also indicated in
Fig. 4.

FIG. 4. The dependence of surface areas and the percentage of micropores to the total surface area for KAM and KAC to the percentages of kaolin
loading (w/w). Closed and open symbols are for KAC and KAM, respectively.
Circles designate micropores and squares total surface area. The percentage of
micropore contribution to the total surface area is also indicated.

97

ADSORBENTS FOR CAROTENE REMOVAL OF RED PALM OIL

TABLE 1
Surface Properties of KAM and KAC
Surface area (m2 /g)

Composition
Activated carbon

BET

Langmuir

Micropore

Micropore volume
(cc/g)

Average pore diameter

(4V/A by BET) (A)

100

20

80

40

60

60

40

80

20

100

1332
(1387)
1013
(829)
746
(640)
439
(640)
247
(368)
19
(37)

1783
(1861)
1366
(1130)
1002
(875)
586
(875)
328
(500)

()

940
(828)
768
(512)
577
(319)
338
(319)
182
(167)
2
(3)

0.4
(0.4)
0.4
(0.2)
0.3
(0.2)
0.2
(0.2)
0.1
(0.1)
0.001
(0.001)

29.4
(28.5)
37.8
(35.8)
49.8
(38.2)
85.1
(38.2)
157.4
(42.0)
232.5
(79.1)

Kaolin

Note. Values in parentheses are of KACs.

A similar trend for the surface areakaolin percentage dependent property was also observed for KAC, but the percentage
contribution of the micropore area to the total surface area has a
different trend. The percentage of micropore increased slightly
from 0 to 20% and then decreased with kaolin loading of >20%.
This is shown in Fig. 4.
An almost linear relationship of surface area to kaolin loading
for KAM indicates that treatment of the mixture at 500 C has no
significant contribution to the surface areas, total or micropore.
The resulting surface areas are merely a simple contribution of
each surface area component of kaolin and activated carbon. Due
to the low micropore surface area of kaolin, its contribution to the
percentage of micropore area of KAM is almost negligible, and
it is therefore dominated by the contribution from the activated
carbon. This is probably why the percentage of micropore is
almost constant up to 80% of kaolin loading.
On the other hand, the resulting total and micropore surface
area of KAC decreased with kaolin loading. Instead of an almost
linear relationship, a slightly downward curve was obtained.
This shows the effect of activationcarbonization of KAC by
2 M sulfuric acid followed by carbonization at 500 C, resulting
in the reduction of the surface area of KAC compared to its
corresponding counterpart, KAM.

the resulting material. This is in parallel with the observation of

the Langmuir Type I isotherm for the material shown in Fig. 2f.
Upon treatment with 2 M sulfuric acid followed by heating
at 500 C, a single pore size distribution of kaolin changed to a
bimodal one with the first strong intensity distribution centered at
and another broad low intensity centered at around
around 100 A

Pore Size Distribution

As a result of desorption isotherms, the BJH desorption pore
size distribution for KAMs and KACs was obtained, and this is
shown in Figs. 5 and 6, respectively.
The BJH pore size distribution for kaolin is presented in
Fig. 5a. As shown in the figure, a strong, high-intensity peak
was observed in its pore size distribucentered at around 400 A
tion. This is in agreement with the Type IV adsorption isotherm
for the mesoporous material shown in Fig. 2a. For activated car pore
bon (Fig. 5f), no peak can be observed in the 10- to 1000-A
diameter region, in agreement with the microporous nature of

FIG. 5. BJH pore size distribution for KAM prepared at various percentages
of kaolin loading (w/w): (a) 100, (b) 80, (c) 60, (d) 40, (e) 20, and (f) 0%.

98

HUSSEIN ET AL.

FIG. 7. Absorption of -carotene on, and its recovery from, KAM and KAC
at various percentages of kaolin loading (w/w). Closed and open symbols are
for KAC and KAM, respectively. Circles are for carotene recovered and squares
are for carotene adsorbed.

the absorbing power of the resulting KAMs and KACs on carotene. These factors were not included in our preliminary
study.
As a result of the adsorption of -carotene on the adsorbents,
we also studied the desorption capability of the -carotene from
the adsorbents. As shown in the figure, desorption is only in
a range of 15% and, in general, desorption decreased as the
loading percentage of kaolin increased for both KAM and KAC.
However, desorption is more favorable as the percentage of activated carbon is higher in the adsorbents.
Peroxide Value
FIG. 6. BJH pore size distribution for KAC prepared at various percentages
of kaolin loading (w/w): (a) 100, (b) 80, (c) 60, (d) 40, (e) 20, and (f) 0%.

This is shown in Fig. 6a. However, no significant changes

500 A.
were observed in the pore size distribution of the activated carbon
treated with 2 M sulfuric acid followed by heating at 500 C, as
shown in Fig. 6f.
Pore size distribution for KAMs and KACs prepared at various
kaolin loading is given in Figs. 5b5e and 6b6e, respectively.
For KAM prepared using 60 and 80% kaolin loading, the pore
size distribution is dominated by a contribution from kaolin. On
the other hand, for KACs, all of them are dominated by the kaolin
property, in parallel with the observation of the dominated kaolin
phase in their corresponding X-ray diffractograms.

Figure 8 shows the peroxide value (PV) of the RPO after

treatment with KAMs and KACs. PV values were used to follow the extent of oxidation of the RPO during the decolorizing
process. Oxidation is likely to occur during RPO in contact with
the adsorbents, KAMs and KACs, and therefore PV gives some
indication of extent of deterioration of the oil.
As shown in Fig. 8, the PV value increases with kaolin loading,
both for KAM and for KAC. However, in general the value for
KAC is always higher than that for KAM. Such an observation
can be related to the acid pretreatment of the precursor to produce
KAC, which somehow contributed to the oxidation of the oil.
Therefore, if this kind of adsorbent is to be used for carotene
removal, the adsorbent prepared without any acid pretreatment

Color Removal of Red Palm Oil

The decoloration power of KAMs and KACs was tested using
RPO and was expressed as the percentage uptake of -carotene.
This is given in Fig. 7. In general, the carotene adsorbed decreased as the kaolin loading increased; however, the optimum
uptake of -carotene was at 20% of kaolin loading for both
KAM and KAC, with about 60 and 70% carotene uptake, respectively. At this point, both the internal and the total surface area
are not at the optimum (see Table 1). This indicates that, apart
from surface area of the adsorbents, other factors such as temperature, adsorbentadsorbate ratio, surface functional groups,
and pH of the adsorbent are also important in contributing to

FIG. 8. Peroxide value of the resulting RPO as a result of carotene adsorption on KAM and KAC at various percentages of kaolin loading (w/w). Closed
and open symbols are for KAC and KAM, respectively.

ADSORBENTS FOR CAROTENE REMOVAL OF RED PALM OIL

99

the adsorbents except the ones prepared using 20 and 40% kaolin
loading. On the other hand, all of the adsorbents show negative
gain for micropore surface area on acid pretreatment, except for
pure kaolin, in which the micropore surface area increased from
2 to 3 m2 /g.
Adsorbents prepared from an activated carbon-rich precursor
show that the micropore surface area is reduced as a result of acid
pretreatment. This is due to a large reduction of micropore area
of pure activated carbon as a result of acid pretreatment, from
940 to 828 m2 /g. Although the micropore area of pure kaolin
has a positive effect on acid pretreatment, a small contribution
of only 3 m2 /g, it results in a negligible contribution to the total
micropore area or the resulting adsorbent.
FIG. 9. Surface area gain (%) as a result of acid pretreatment of the adsorbents prepared at various percentages of kaolin loading (w/w). Closed and open
symbols are for total and micropore surface area, respectively.

at our experimental condition is superior, in which it introduced

less degradation of the resulting oil.
The Effect of Sulfuric Acid Pretreatment
In order to observe the effect of sulfuric acid pretreatment, we
monitored the physicochemical properties of the resulting adsorbents as well as the amount of carotene adsorbed and desorbed
and the property of RPO after treatment with the adsorbents,
in particular its PV value. The resulting properties were then
compared to the corresponding properties of the untreated adsorbents or RPO, which was treated with the adsorbents prepared
without any acid pretreatment.

Adsorption and Desorption of Carotene from RPO

and Its Effect on PV Value
Apart from physicochemical properties, we also analyzed the
effect of sulfuric acid pretreatment on the resulting properties
of the adsorbents toward the adsorption and desorption of carotene from RPO. In addition, the effect on PV value, which
reflects the extent of oil oxidation, was also evaluated. This is
shown in Fig. 10, in which the effect was calculated as
Gain (%) = (Vp Va ) 100/Va ,

[3]

where Vp and Va are the values for either carotene adsorbed

or carotene recovered from the adsorbent prepared with and
without acid pretreatment, respectively. Similarly, Vp and Va are

Physicochemical Property of the Adsorbents

Treatment of kaolin with sulfuric acid followed by heating
at 500 C resulted in well-defined crystallinity of the adsorbent.
This is clear by comparison of Figs. 1a and 1b for 100% kaolin
loading for KAM and KAC, respectively. In addition, the basal
was maintained, which means that
spacing of around 7.14 A
the layered structure of the pure clay was not collapsed. As
was discussed earlier, acid pretreatment of the kaolincarbon
mixture resulting in adsorbents with enhanced crystallinity, even
at very low kaolin loading, 20%, with no alteration of the basal
spacing of the kaolinite phase, can be observed, as shown in
Fig. 1b.
Figure 9 shows the effect of sulfuric acid pretreatment on the
total and micropore surface areas of the resulting adsorbents.
The surface area gain was calculated as
Surface area gain (%) = (Sp Sa ) 100/Sa ,

[2]

where Sp and Sa are the surface area (total or micropore) of the

adsorbent with and without acid pretreatment, respectively.
A positive total surface area gain, i.e., the total surface area increased for the adsorbent prepared with acid pretreatment compared to that without acid pretreatment, was observed for all of

FIG. 10. Comparison of the amount of carotene adsorbed and recovered

from adsorbents prepared with and without acid pretreatment at various percentages of kaolin loading (w/w). Closed and open symbols are for carotene
adsorbed and recovered, respectively. The PV value of the resulting treated
RPO is also indicated (striped bars).

100

HUSSEIN ET AL.

also used for the PV value of resulting RPO in contact with the
adsorbents prepared with and without acid pretreatment, respectively.
For various kaolin loadings, the percentage of carotene adsorbed was found to be improved with the acid pretreatment
of the adsorbent prior the activationcarbonization process, as
shown in Fig. 7. Figure 10 compares the performance of adsorbents on the acid pretreatment to those without pretreatment.
Carotene adsorption increased in a range of 315%, and the maximum was found for the adsorbent prepared using 40% kaolin
loading. However, the same treatment hindered the ability of the
adsorbed carotene to be desorbed from the adsorbents if 4080%
kaolin loading was used. On the other hand, such a treatment
has a positive effect, i.e., it enhanced the desorption of carotene
if pure activated carbon, kaolin, or a mixture of both with 20%
kaolin loading was used to prepare the adsorbents. Pure kaolin or
activated carbon improved its carotene desorption by acid pretreatment, about 49 and 47%, respectively. Although in terms
of percentage, the increase is quite substantial, the actual value
is only 4.7 and 3.2% removal of carotene for acid pretreated
and untreated, respectively, for activated carbon, compared to
only 1.1 and 0.75%, respectively, for kaolin. This is shown in
Fig. 7.
Except for the adsorbent prepared using 20% kaolin loading,
acid pretreatment of the adsorbent resulting in a higher PV value
of the RPO treated with it by about 25%, as shown in Fig. 10. In
general, this shows that acid pretreatment, produced adsorbents
when in contact with RPO, oxidized it, and hence increased the
PV value.
CONCLUSIONS

Kaolincarbon adsorbents can be prepared with and without

sulfuric acid treatment prior to the activationcarbonization process at medium temperature, 500 C. The surface area and the
porosity nature of the resulting materials were found to be related to the kaolin-loading percentage, and an optimum surface
area was obtained at 20% kaolin loading.
Sulfuric acid pretreatment of the precursor prior to the
carbonizationactivation processes resulted in enhancement of
the total surface area but reduced the micropore component of
the resulting adsorbents. At the same time, this improved the
carotene adsorption capacity from red palm oil. However, recovery of carotene from the carotene-adsorbed adsorbent is only
improved when the acid pretreatment of the precursor was done

at a high loading percentage of activated carbon. Similarly, the

peroxide value increases.
In general, the surface area of the activated kaolincarbon
mixture and its complexes decreased with the kaolin-loading
percentage. However, the optimum carotene absorption was obtained at 20% kaolin loading for both adsorbents, with about
60 and 70% carotene removal, respectively. A further increase
in the kaolin loading resulted in the decrease of the carotene
removal. About 34% of the carotene adsorbed can be recovered
from both types of adsorbents at the optimum uptake in which
the percentage recovered decreased with the increase of kaolin
loading. On the other hand, the peroxide value increased with
kaolin loading at around 5464 mEq/kg for both materials.
It was also found that carotene uptake by the adsorbents is
high if the adsorbent contains a high percentage of activated
carbon. Similarly, carotene recovery is high and less oxidation
can be observed, as indicated by the lower PV value. Therefore,
for the adsorbents prepared in this work, a higher content of
activated carbon is more desirable, as it give better adsorption
and desorption performance and fewer oxidation effects to the
resulting oil, RPO.
ACKNOWLEDGMENT
The support of this research by the Ministry of Science, Technology, and Environment, Malaysia, under IRPA Mechanism, Grant 09-02-04-0027, is gratefully
acknowledged.

REFERENCES
1. Hussein, M. Z., Tarmizi, R. S. H., Zainal, Z., Ibrahim, R., and Badri, M.,
Carbon 34, 1447 (1996).
2. Hussein, M. Z., Zainal, Z., Ibrahim, R., Kheong, K. K., and Badri, M.,
J. Chem. Tech. Biotechnol. 64, 35 (1995).
3. Rodriguez-Reinoso, F., Lope-Gonzalez, J. de D., and Berenguer, C., Carbon
20, 513 (1982).
4. Lope-Gonzalez, J. de D., Marteninez-Vilchez, F., and Rodriguez-Reinoso,
F., Carbon 18, 413 (1980).
5. Pollard, S. J. T., Sollars, and Perry, R, J. Chem. Technol. Biotechnol. 50, 265
(1991).
6. Lee, C. K., Low, K. S., and Chung, L. C., J. Chem. Tech. Biotechnol. 69, 93
(1997).
7. Gee, I. L., Sollars, C. J., Fowler, G., Ouki, S. K., and Perry, R., J. Chem.
Technol. Biotechnol. 72, 329 (1998).
8. Barret, E. P., Joyner, L. S., and Halenda, P. P., J. Am. Chem. Soc. 73, 373
(1951).
9. Lin, S. W., Tang, T. S., and Tan, T. A., in PORIM Test Methods: Methods
of Test for Palm Oil and Palm Oil Products, PORIM Publication, Kuala
Lumpur, 1995.