LWT - Food Science and Technology 59 (2014) 1258e1264

Contents lists available at ScienceDirect

LWT - Food Science and Technology

journal homepage: [Link]/locate/lwt

Effect of type of bleaching earth on the final color of refined palm oil

c,
Simone M. Silva a, Klicia A. Sampaio c, Roberta Ceriani b, Roland Verhe
c d a, *
Christian Stevens , Wim De Greyt , Antonio J.A. Meirelles
a
EXTRAE, Department of Food Engineering, School of Food Engineering, UNICAMP, Campinas, SP, Brazil
b
Department of Processes and Products, School of Chemical Engineering, UNICAMP, Campinas, SP, Brazil
c
Department of Sustainable Organic Chemistry and Technology, Faculty of Bioscience Engineering, University of Ghent, Ghent, Belgium
d
Desmet Ballestra, Zaventem, Belgium

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

Article history: Although studies indicate chemical changes during bleaching such as carotene and unsaturated fatty
Received 29 November 2013 acids oxidation, which are probably responsible for the color fixation of palm oil, this process is not very
Received in revised form clear. The objective of this study was to investigate the effect of type and amount of bleaching earth (BE)
8 May 2014
on the final quality of refined palm oils, especially on the oxidative state and color. Two types of
Accepted 12 May 2014
bleaching earth were tested, one natural (NBE) and one acid-activated (ABE) (0.5e3.0% w/w). Crude palm
Available online 27 May 2014
oils were bleached at 105
C, during 30 min at 50 mmHg pressure. Afterwards, a deodorization step was
performed at 260
C, 3 mbar, 1.5% steam during 60 min. These refining procedures were evaluated after
Keywords:
Oxidation mechanism
each step by measuring b-carotene, color, peroxide (PV) and p-anisidine (pAV) values. It was observed
b-carotene that both BE can decompose peroxides. However, a maximum pAV followed by a decrease was observed
Deodorization for ABE while the pAV remains approximately constant at a maximum for NBE, suggesting only ABE
Color fixation catalytically decomposes secondary oxidation products. The color after deodorization was inversely
Edible oil proportional to pAV when bleaching was performed with ABE, even though the oil has a lighter color
after deodorization.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction In the case of palm oil refining, bleaching is integrated in a so-

called dry pretreatment step which aims at simultaneous
Crude palm oil is produced from palm fruits and contains about bleaching and element removal with minimal formation of
1% of minor components, such as phospholipids, metals, phytos- oxidation products. Complete removal of carotenes during
terols, carotenes, tocopherols and tocotrienols. Its high carotenoid bleaching is not possible and also not necessary since these
content (500e700 mg/kg) is responsible for the dark orange color components are not heat stable and can be degraded during high
(Gee, 2007). Crude palm oil has to be refined to achieve desirable temperature deodorization (Gibon et al., 2007). In fact, some
characteristics, such as a light color, bland taste and a good oxida- refineries only remove 20% of the color of the carotenoid content
tive stability (Gibon, De Greyt, & Kellens, 2007; Sampaio, Ceriani, of the crude palm oil at the bleaching step, and after 20 min at
Silva, Taham, & Meirelles, 2011). Physical refining is generally 240
C more than 98% of the total carotenoids has been destroyed
preferred to avoid a too high neutral oil loss (Rossi, Gianazza, (Maclellan, 1983). Part of the color pigments are physically
Alamprese, & Stanga, 2001), and usually consists of a bleaching adsorbed by bleaching earth. Other components are chemically
and a deodorization/deacidification step (Gibon et al., 2007). Both bound to bleaching clay via covalent or ionic bonds (Gibon et al.,
steps are important for the removal of coloring components. Part 2007; Taylor, 2005). Moreover, bleaching earth acidity is related
of the color pigments are removed during bleaching by adsorption to the pigment adsorption capacity (Taylor, 2005). Acid activated
on a suitable bleaching earth. The remaining color components bleaching earth work simultaneously as a solid, acidic catalyst,
are then thermally degraded during deodorization at high adsorptive agent, cation exchanger and filter aid (Zschau & Grp,
temperature. 2001), whereas neutral bleaching earth acts basically only as an
adsorptive material (Taylor, 2005). In fact, there is no consensus
on the adsorption mechanisms of b-carotene onto bleaching
* Corresponding author.
earth. Boki, Kubo, Wada, & Tamura (1992) found that pigment
E-mail address: tomze@[Link] (A.J.A. Meirelles). adsorption onto clay minerals occurs through physisorption.

[Link]
0023-6438/© 2014 Elsevier Ltd. All rights reserved.
 S.M. Silva et al. / LWT - Food Science and Technology 59 (2014) 1258e1264 1259

Sabah, Cinar, & Celik (2007) suggested that b-carotene adsorp- Table 1
tion onto sepiolite is also physical. On the other hand, according Palm Oil Color, b-carotenes and elements content (mg/kg) after bleaching and
deodorization using acid activated (ABE) and neutral (NBE) bleaching earths.
to Khoo, Morsingh, & Liew (1979), b-carotene adsorption onto
different commercial bleaching earths is not purely physical. ABE NBE
Later, other works found that b-carotene adsorption onto acid 0.5 1.0 3.0 0.5 1.0 3.0
activated bleaching earth (Sarier & Guler, 1988) and onto silica
After Bleaching
gel (Ahmad, Chan, Shukor, & Mashitah, 2009) was controlled by Elements content
chemisorption. (mg/kg)
It is well accepted that there is no correlation between the b- P 2.7 ± 0.1 0.5 ± 0.1 0.2 ± 0.1 3.1 ± 0.1 0.6 ± 0.1 0.2 ± 0.0
Fe 0.7 ± 0.1 0.2 ± 0.0 0.2 ± 0.0 0.5 ± 0.0 0.1 ± 0.0 0.2 ± 0.0
carotene content after bleaching and the palm oil final color
Carotenes 258 83 10 280 87 NDa
(Dijkstra & Segers, 2007; Fraser & Frankl, 1981; Maclellan, 1983). (mg/kg)b
Some studies suggest that in the finished oil, the color is mostly due Color (5 ¼ in) 68.0 42.0 15.8 68.0 39 12.5
to high molecular weight (HMW) compounds derived from Red (R)
oxidation reactions, especially in the case of carotenoids (Fraser & After deodorization
Color (5 ¼ in) 8.2 7.3 3.4 9.1 8.4 6.1
Frankl, 1981). Also, aggressive bleaching conditions, as excessive
Red (R)
temperatures or time process or both, lead to darker final colors
a
ND: Not detected.
(Maclellan, 1983). b
Carotenes were not detected in any sample after deodorization.
Recently, there are several research groups studying b-caro-
tene oxidation in lipid/organic systems (Benevides, Veloso,
Pereira, & de Andrade, 2011; Bosser, Paplorey, & Belin, 1995;
Rodriguez & Rodriguez-Amaya, 2007; Wache, Bosser-DeRatuld, absorbance at 446 nm
DOBI ¼ (1)
Lhuguenot, & Belin, 2003; Zeb, 2012; Zeb & Murkovic, 2013a, absorbance at 268 nm
2013b) concerning b-carotene antioxidant activity, off-flavor
The higher the DOBI value, the lower the amount of oxidized
compounds formation and synthesis. However, there is a lack
carotenes, the more efficient it will be to heat bleach the oil, as color
in the literature in works concerning the b-carotene pathways
due to oxidized carotenes is difficult to remove. DOBI values be-
during refining practices and its importance in the quality of fully
tween 2.5 and 4.0 indicate average to good crude oil quality. Values
refined oils. In this context, the goal of this work was to study the
below 2.0 indicate a poor quality crude palm oil which is difficult to
effect of the type of bleaching earth used during bleaching on the
bleach.
color of fully refined palm oil. Moreover, a hypothesis is proposed
to explain how the type of bleaching earth can affect the
2.1.4. Carotenes content
oxidation pathways.
Carotenes content, expressed as b-carotene, was determined
measuring the absorbances at 446 nm of samples homogenized
2. Material and methods
and diluted in iso-octane, according to PORIM (1990).
2.1. Oil characterization
2.1.5. p-Anisidine and peroxide values
Standard quality parameters, including free fatty acids (FFA) Peroxide (PV) and p-anisidine (p-AnV) values were determined
content, color, carotene content, deterioration of bleachability in- according to the AOCS Official Methods Cd 8-53 and Cd 18-90,
dex (DOBI) and element content, were analyzed for crude, bleached respectively (AOCS, 1998).
and full refined palm oil samples.
2.1.6. Elements content
2.1.1. Color The content of Phosphorus (P) and Iron (Fe) were measured
The sample color was determined according to the AOCS Official using an Inductively Coupled Plasma (ICP) method, according to the
Method Cc 13e-92, using a Lovibond Tintometer Color Scale at AOCS official method Ca 20-99, using kerosene as solvent (AOCS,
70
C, ensuring that color was determined in a completely 1998).
melted sample. Analyses were carried out using 500 ¼ (133.4 mm)
glass cells. In the cases of crude palm oil samples, the color was 2.2. Refining procedure
darker than scale in a 500 ¼ glass cell and a 100 (25.4 mm) was used
(AOCS, 1998). Bleaching trials were performed according to the following
steps: heating the crude palm oil to 85
C; adding citric acid (0.3% of
2.1.2. Free Fatty Acids (FFA) a 30% solution); high shear mixing (16,000 rpm during 1 min);
Free Fatty Acids content (FFA) was determined according to the addition of Bleaching Earth; 15 min of maturation time at 85
C and
AOCS Official Method Ca 5-40, by titration. Results were expressed atmospheric pressure; applying vacuum (50 mbar) and maintain-
as percentage of palmitic acid (C16:0, MW ¼ 256 g/mol) ing bleaching conditions for 30 min at 105e110
C and removing
(AOCS, 1998). the bleaching earth by filtration over a Buchner filter (pore size
11 mm, Whatman).
2.1.3. Deterioration of the bleachability index (DOBI) Experiments were performed using two kinds of bleaching
The deterioration of bleachability index (DOBI) was measured earth: non-activated bleaching earth (Pure Flo B80, purchased from
using a UVevis spectrophotometer. This index corresponds to the Oil-Dri, here so-called NBE) and acid activated bleaching earth
ratio of the spectrophotometric absorbance at 446 nm (non- (Tonsil OPT. 210 FF, purchased from Süd-Chemie AG, here so-called
oxidized carotenes maximum) to the absorbance at 268 nm ABE).
(oxidized carotenes maximum). After bleaching, the oil was submitted to a deodorization step
About 0.1 mg of completely molten and homogenized oil is since some color effects can just be observed after heat bleaching.
diluted in 25 mL of iso-octane. Absorbencies are measured at To avoid differences due to storage effects, the deodorization was
268 nm and 446 nm. The DOBI value is obtained by Eq. (1). always carried out one day after the bleaching. The deodorization
1260 S.M. Silva et al. / LWT - Food Science and Technology 59 (2014) 1258e1264

was performed in a lab-scale equipment described by Petrauskaite, decrease back to the initial value. In the case of NBE, p-AnV
De Greyt, and Kellens (2000). Deodorization parameters were fixed shows an increasing trend up to an approximately constant
to all experiments and set as 260
C of deodorization temperature, value. It is interesting to note that maximum p-AnV values
3.0 mbar, 1.5% steam and 60 min of residence time. coincide with the point the peroxides value reaches zero for the
first time, suggesting that the peroxides are being converted into
3. Results secondary oxidation products.
Different p-AnV values were obtained by adjusting the bleach-
Crude palm oil (obtained from a local processor) was charac- ing earth amount and by repeating the bleaching procedure in fully
terized regarding FFA, DOBI, color, carotenes and elements content refined oils. A correlation between p-AnV after bleaching and palm
(mg/kg). The DOBI value was 2.06 ± 0.01, indicating that the final oil color after deodorization was observed when ABE was used
color is difficult to predict after complete refining of the oil (Gibon (Fig. 2). The same correlation was not observed for NBE.
et al., 2007). The initial content (mg/kg) of phosphorus, iron and As a second part of this work, the effect of the amount of citric
carotenes was 19.1 ± 0.2, 7.4 ± 0.9 and 467 ± 2, respectively. The acid added during bleaching was evaluated. Citric acid can act in
crude palm oil presented an initial FFA of 4.61%, expressed in pal- two ways during oil degumming: firstly, displacing phosphatidic
mitic acid, and 37.8 h OSI. acid (PA) in its salts; secondly, forming a chelating agent capable to
A bleaching procedure was tested using both kinds of bleaching form a stronger bond with alkaline earth than PA does (Dijkstra,
earth at three different concentrations: 0.5, 1.5 and 3.0% (Table 1). 2010). Here, we assume that by adding more citric acid, more ele-
Phosphorus and iron contents lower than 3.0 and 0.3 mg/kg, ments species can be chelated, reducing their concentration in the
respectively, were achieved by both bleaching earths when using oil. Moreover, citric acid reduces the pH of the solution and com-
3.0%. ABE leads to a better reduction of phosphorus especially when petes with compounds to be adsorbed. The effect of the citric acid
small amounts of bleaching earth are used. However, increasing the (added as 30% solution) was tested using 2.0% of bleaching earth
amount, this effect becomes less important, and the phosphorus (Table 2). Note that this represents an increase in citric acid and
content obtained using each bleaching earth is quite similar. The water addition.
opposite behavior was observed for iron adsorption. The oils showed a higher FFA after bleaching with ABE than
Concerning the carotenes analysis, it can be noted that for the those bleached with NBE. This can be due to the inherent acidity of
lower concentration level tested, ABE resulted in a higher adsorp- ABE or due to a catalytic effect forming FFA. The phosphorus
tion of carotene than NBE. Using larger quantities, NBE gives a adsorption decreases with the amount of added citric acid for both
higher adsorption. It was not possible to observe a relation between bleaching earths, but the decrease is more significant with NBE.
the carotene content after bleaching and the color after deodor- There is no difference between the final Iron content using the
ization (Table 1), as suggested by Taylor (2005). same bleaching earth.
Regarding the color, it can be noted that after bleaching using Concerning the carotenes content, increasing the citric acid
0.5% led to 68.0 R for both BE. When a higher amount of BE was concentration resulted in different behavior for each bleaching
used, differences in color could be observed; NBE leads to an oil of earth. In the case of ABE; adding more citric acid caused that more
about 3.0 R lighter then ABE after bleaching. After deodorization carotenes were adsorbed. On the other hand, when NBE was used
the oil bleached with ABE presents a lighter color, even though it as bleaching earth, the increasing amount of citric acid resulted in a
was darker after bleaching. It is interesting to highlight that the lower adsorption of carotenes. It was not possible to observe any
carotene content and Lovibond color obtained after bleaching with relation between the color after bleaching and after deodorization,
ABE were higher than that obtained with NBE, however, it gave as suggested by Fraser & Frankl (1981). In the case of ABE, the use of
instead a lighter color after deodorization. more citric acid leads to a lower color after bleaching but a higher
The oxidative state of the oils after bleaching using ABE and color after deodorization. On the other hand, when NBE was used,
NBE are presented in Fig. 1. For both bleaching earths, peroxides more citric acid leads to a higher color after bleaching and a lower
value decrease with BE concentration. Note that, amounts color after deodorization. From those results, it can be inferred that
slightly higher than 1.0% ABE reduces the PV to zero, while for the chelating power is not the factor affecting the palm oil final
the NBE about 2.0% is necessary. Small amounts of BE increases color. A possible explanation is that the pH interferes in b-carotenes
the p-AnV. In the case of ABE, p-AnV increases and then starts to oxidation pathways, and this will be further discussed in the
following section.

Fig. 1. Oxidative State of Bleached Oils: Relative Peroxide and p-Anisidine values
(value/initial value) vs amount of bleaching earth (% w/w). (Legend: Peroxide value, - Fig. 2. p-Anisidine value vs Colour in oils bleached with different amounts of
ABE , NBE/p-Anisidine value, A ABE ◊ NBE). bleaching earth (Legend: - ABE , NBE).
 S.M. Silva et al. / LWT - Food Science and Technology 59 (2014) 1258e1264 1261

Table 2 radicals undergo addition of oxygen producing dicarbonyls (Kamal-

Effect of citric acid and water amount during bleaching using 2.0% acid activated Eldin, 2003). It is clear, that allylic hydrogen abstraction does not
(ABE) and neutral (NBE) bleaching earths.
occurs in bleaching conditions due to its moderate temperature and
ABE NBE pressures, being always lower than 50 mbar. Moreover, b-carotene
0.09% CA 0.27% CA 0.09% CA 0.27% CA behaves differently in different oils and under different conditions,
i.e. temperature and lipid system composition (Zeb & Murkovic,
0.21% water 0.63% water 0.21% water 0.63% water
2013b).
After Bleaching
Burton and Ingold (1984) suggested that b-carotene reacts
FFA (%) 5.0 5.0 4.7 4.8
Elements content with peroxyl groups by addition, rather than by hydrogen
(mg/kg) abstraction. Liebler & McClure (1996) identified radical adducts
P 0.4 ± 0.1 0.5 ± 0.1 0.6 ± 0.0 0.8 ± 0.1 formed during b-carotene oxidation, which then combines
Fe 0.2 ± 0.0 0.2 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 with a second radical to form an addition product. The formation
Carotenes (mg/kg)a 40 21 13 19
Color (5 ¼ in) 29.0 25.0 19.2 22.0
of alkyl- and alkoxyl-containing addition products indicates
Red (R) that both may add directly to b-carotene. In contrast, peroxyl
After deodorization radical addition yields an unstable intermediate radical
Color (5 ¼ in) 6.1 7.3 8.9 8.2 adduct that collapses to an epoxide and releases an alkoxyl
Red (R)
radical.
a
Carotenes were not detected in any sample after deodorization. The polarity of the medium determines the pathways that b-
carotene oxidation undergoes: in nonpolar solvents, only addition
radicals are formed, meanwhile in polar ones carotenoid radical
4. Discussion cations are formed. For both solvents, the first product is an addi-
tion radical formed between the acylperoxyl radical and the
In this work, no correlation was found between the b-carotene carotenoid (Eq. (2), Fig. 3) (El-Agamey & McGarvey, 2003). In fact,
concentration of crude oil and the final color of the palm oil. On the the Gibbs free energy presents similar negative values for both
other hand, a high correlation was found between p-AnV after system polarities, thus, the exergonicity of reactions depends
bleaching with ABE and the palm oil color after deodorization/heat mostly on the nature of the free radical, being OH radical the most
bleaching. The same correlation was not observed when the exergonic and peroxyl radicals the least one (Martinez, Vargas, &
bleaching was done with NBE. It is also interesting to highlight that Galano, 2010). Those radicals have no reactivity toward oxygen,
heat bleaching was more efficient after bleaching with ABE, even even at high oxygen pressures as 760 mmHg (Eq. (3)) (El-Agamey &
though NBE was capable of removing more efficiently the color McGarvey, 2003).
during bleaching. Oxidation of compounds such as b-carotenes may
be an explanation for those results as described hereafter. BC þ ROO / ROOBC (2)
Tonsil OPT 210 FF is an acid activated bleaching earth manu-
factured by acid activation of calcium bentonite. It has acidic and ROOBC þ O2 / ✕ (3)
catalytic activity, leading most importantly to hydroperoxide
decomposition, forming sub-products such as aldehydes, ketones Subsequent steps of b-carotene oxidation follow different
and conjugated polyenes (Zschau & Grp, 2001). Pure Flo B 80 is a pathways for polar and non-polar solvents (Fig. 4). In non-polar
neutral bleaching earth with no catalytic activity. These properties solvents, the radicals decompose in epoxides and cyclic ethers,
can be confirmed through oxidative state data (Fig. 1). Note, for therefore, with an acyloxyl radical elimination (Eq. (4)) (El-Agamey
instance, a smaller amount of ABE is needed to reach zero peroxides & McGarvey, 2003), leading to volatile products such as apocar-
compared to NBE. Furthermore, the first point presenting a otenals (Krinsky & Yeum, 2003). Note that this reaction releases a
maximum p-AnV value (secondary oxidation products) is corre- radical, thus, showing no net consumption of radical, being an
lated to the minimum PV value. Note also that ABE decreases p-AnV autoxidation reaction (Liebler, 1993).
until a constant value and NBE keeps it constant at the maximum.
Indeed, as suggested by Sarier & Guler (1988), b-carotene which ROOBC þ ROO / “nonradical product” (4)
remains in solution with acid activated bleaching earth is rapidly
oxidized later on, even more than those submitted to oxygen for Nonetheless, in polar solvents, an ion-pair composed by a per-
48 h. Consequently, it may be said that acid activated bleaching oxyl anion (ROO) and a cation (CARþ) is formed, which absorbs
earth initiates the oxidation of unadsorbed b-carotene. near-infrared radiation (El-Agamey & McGarvey, 2003) and after, it
Carotenoids can react with radical species in three different is transformed to carotenoid radical cation. Even though vegetable
ways, including (1) radical addition, (2) electron transfer to the oil is nonpolar, BE negative sites may make a polar layer close to its
radical or (3) allylic hydrogen abstraction (Bonnie & Choo, 1999).
Which mechanism will take place depends on the reaction condi-
tions. For instance, by increasing the oxygen concentration, the
formation of secondary peroxyl radicals becomes important,
resulting in the loss of antioxidant behavior. With a sufficiently
high oxygen partial pressures (above 150 mmHg) (Burton & Ingold,
1984), b-carotene reactions have a pro-oxidant effect since they
could generate more radicals than they consume (Krinsky & Yeum,
2003). Electron transfer reactions have been reported either in
formation of a carotenoid cation CARþ, anion CAR or in the for-
mation of an alkyl radical (Krinsky & Yeum, 2003), stable because of
its resonance structure (Kamal-Eldin, 2003). Finally, allylic
hydrogen abstraction by peroxyl radical occurs at an oxygen pres-
sure of less than 760 mmHg, producing a carotene radical. Those Fig. 3. First Addition Radical formed in both polar and non polar media.
1262 S.M. Silva et al. / LWT - Food Science and Technology 59 (2014) 1258e1264

Fig. 4. b-carotene oxidation pathways using acid activated (ABE) and neutral (NBE) bleaching earths.

surface, supporting reactions associated with polar environments. increasing competition among compounds to be adsorbed and
Negative charged sites of ABE can readily adsorb those cations. reducing the solution pH. The increase of competition among
Indeed, cationic compounds are preferentially adsorbed by acidic compounds to be adsorbed has a negative effect, as it reduces the
centers of bleaching earth, as suggested by Zschau & Grp (2001), adsorption of P and Fe. This phenomenon is observed in
due to a chemisorption mechanism instead of physisorption (Kent,
2008). According to Silva et al. (2013), b-carotenes are chemisorbed
by ABE.
During the deodorization step, b-carotene degradation prog-
ress follows different mechanisms. Heat excited carotenoids un-
dergo cisetrans isomerization or react with oxygen to form
diradicals (Kamal-Eldin, 2003). According to Rodriguez &
Rodriguez-Amaya (2007) both cis and trans carotenes are sus-
ceptible to oxidation. b-carotene can undergo autoxidation much
easier than compounds containing fewer conjugated double
bonds, as b-apo-80 -carotenal and retinal, due to transecis thermal
isomerization. During twisting of the b-carotene backbone un-
paired spin density will develop in each half of the molecule,
reaching a maximum in the perpendicular transition state and
forming cyclic peroxides which may self-initiate autoxidation
reactions. The main stable products to thermal degradation are
higher molecular weight components, epoxy-b-carotenes, apo-
carotenals and apo-carotenones, other low molecular weight
di- and tri-oxygenated compounds, probably including carbox-
ylic and/or peracids, and carbon dioxide (Mordi et al., 1993)
(Fig. 5). Marty and Berset (1990) suggested b-carotene degrades
by progressive shortening of the polyene chain by two carbon
atoms at a time. So, oxidative break of a double bond in the b-
carotene molecule leads to the formation of two carbonyl frag-
ments of which one may be colorless. Benevides et al. (2011)
studied b-carotene oxidation by ozone in organic model and
demonstrated that the oxidized isomers formed depends on the
oxidation agent.
The effect of citric acid addition on the palm oil final color
reinforces this hypothesis. The citric acid can induce an Fig. 5. Products formed during b-carotene oxidation through radical addition.
 S.M. Silva et al. / LWT - Food Science and Technology 59 (2014) 1258e1264 1263

processes using both BE, and the concentrations of those ele- Gee, P. T. (2007). Analytical characteristics of crude and refined palm oil and frac-
tions. European Journal of Lipid Science and Technology, 109(4), 373e379. http://
ments after processing with more citric acid are higher. The
[Link]/10.1002/ejlt.200600264.
reduction of the pH can change the b-carotene pathways, as * Gibon, V., De Greyt, W., & Kellens, M. (2007). Palm oil refining. European Journal of
explained previously. In the case of the process using NBE, the Lipid Science and Technology, 109(4), 315e335. [Link]
use of more citric acid leads to a more polar environment and ejlt.200600307 (This paper presents a complete review on palm oil refining pro-
cedures, including its main problems such as the color fixation discussed in our
makes the formation of cations more favorable. The process using work. In fact, this work highlights about the problem of color fixation for the palm
ABE shows an already sufficient low pH, so that even with less oil refiners.).
citric acid those reactions were promoted. * Kamal-Eldin, A. (2003). Lipid oxidation pathways. Chaimpaing, IL: AOCS Press (This
book presents three possible pathways for b-carotene oxidation and how different
In this way, the positive effect of the pH is more pronounced environments can affect it. It was important to the authors to conclude that the
with NBE, whilst the negative effect of competition with ABE. environment modification made by the bleaching earth could affect the oxidation
Processing with NBE tends to have the same behavior as processing pathway.).
Kent, K. (2008). Adsorption. In Albright's chemical engineering handbook (pp.
with ABE when more citric acid is added: lower carotenes 1119e1171). CRC Press.
adsorption, darker color after bleaching and lighter color after Khoo, L. E., Morsingh, F., & Liew, K. Y. (1979). Adsorption of beta-carotene-1 by
deodorization. bleaching earths. Journal of the American Oil Chemists Society, 56(7), 672e675.
[Link]
This work presented important results to improve the un- * Krinsky, N. I., & Yeum, K. J. (2003). Carotenoideradical interactions. Biochemical
derstanding of palm oil refining and some concluding remarks and Biophysical Research Communications, 305(3), 754e760. [Link]
can be drawn so far: the type of bleaching earth can define 10.1016/S0006-291x(03)00816-7 (This reference shows that b-carotene can react
through cation, forming an alkyl radical. This is the mechanism we ascribe to
oxidation pathways in two different manners. At the first
oxidation when using acid activated bleaching earth.).
moment, ABE decomposes peroxides in alkoxyl and alkyl radi- Liebler, D. C. (1993). Antioxidant reactions of carotenoids. Carotenoids in Human
cals, which can react faster with b-carotene. Later, ABE creates a Health, 691, 20e31. [Link]
more polar environment, supporting radical cation formation Liebler, D. C., & McClure, T. D. (1996). Antioxidant reactions of beta-carotene:
identification of carotenoid-radical adducts. Chemical Research in Toxicology,
which absorbs near infra-red. Furthermore, acid activation en- 9(1), 8e11. [Link]
hances the adsorptive power due to an increasing surface area Maclellan, M. (1983). Palm oil. Journal of the American Oil Chemists Society, 60(2),
(Taylor, 2005). 368e373. [Link]
Martinez, A., Vargas, R., & Galano, A. (2010). Theoretical study on the chemical fate
of adducts formed through free radical addition reactions to carotenoids.
Theoretical Chemistry Accounts, 127(5e6), 595e603. [Link]
Acknowledgments s00214-010-0753-3.
Marty, C., & Berset, C. (1990). Factors affecting the thermal-degradation of all-trans-
The authors would like to thank Desmet Ballestra R&D Centre beta-carotene. Journal of Agricultural and Food Chemistry, 38(4), 1063e1067.
[Link]
for providing all facilities necessary for the development of this * Mordi, R. C., Walton, J. C., Burton, G. W., Hughes, L., Ingold, K. U., Lindsay, D. A.,
work. R. Ceriani and A.J.A. Meirelles acknowledge Sa ~o Paulo et al. (1993). Oxidative-degradation of beta-carotene and beta-apo-8'-carotenal.
Research Foundation (2008/56258-8; 2010/16634-0) and CNPq Tetrahedron, 49(4), 911e928. [Link]
(This reference presents the main stable products of b-carotene, such as high
(306250/2007-1; 301999/2010-4) for the financial support. S.M. molecular weight components (epoxy-b-carotenes, apo-carotenals and apo-
Silva acknowledges CNPq (140283/2009-9) and Coordenaç~ ao de carotenones).).
Aperfeiçoamento de Pessoal de Nível Superior (0099-11-2) for Petrauskaite, V., De Greyt, W. F., & Kellens, M. J. (2000). Physical refining of coconut
oil: effect of crude oil quality and deodorization conditions on neutral oil loss.
scholarships. Journal of the American Oil Chemists Society, 77(6), 581e586. [Link]
10.1007/s11746-000-0093-6.
PORIM. (1990). Test methods carotene content.
References Rodriguez, E. B., & Rodriguez-Amaya, D. B. (2007). Formation of apocarotenals and
epoxycarotenoids from beta-carotene by chemical reactions and by autoxida-
Ahmad, A. L., Chan, C. Y., Shukor, S. R. A., & Mashitah, M. D. (2009). Adsorption tion in model systems and processed foods. Food Chemistry, 101(2), 563e572.
kinetics and thermodynamics of beta-carotene on silica-based adsorbent. [Link]
Chemical Engineering Journal, 148(2e3), 378e384. [Link] Rossi, M., Gianazza, M., Alamprese, C., & Stanga, F. (2001). The effect of bleaching
[Link].2008.09.011. and physical refining on color and minor components of palm oil. Journal of the
AOCS. (1998). Methods and recommended practices of the American Oil Chemists' American Oil Chemists Society, 78(10), 1051e1055. [Link]
Society. Champaign. s11746-001-0387-8.
Benevides, C. M. D., Veloso, M. C. D., Pereira, P. A. D., & de Andrade, J. B. (2011). Sabah, E., Cinar, M., & Celik, M. S. (2007). Decolorization of vegetable oils:
A chemical study of beta-carotene oxidation by ozone in an organic model adsorption mechanism of beta-carotene on acid-activated sepiolite. Food
system and the identification of the resulting products. Food Chemistry, 126(3), Chemistry, 100(4), 1661e1668. [Link]
927e934. [Link] 05.12.052.
Boki, K., Kubo, M., Wada, T., & Tamura, T. (1992). Bleaching of alkali-refined vege- Sampaio, K. A., Ceriani, R., Silva, S. M., Taham, T., & Meirelles, A. J. A. (2011). Steam
table-oils with clay-minerals. Journal of the American Oil Chemists Society, 69(3), deacidification of palm oil. Food and Bioproducts Processing, 89(C4), 383e390.
232e236. [Link] [Link]
Bonnie, T. Y. P., & Choo, Y. M. (1999). Oxidation and thermal degradation of carot- * Sarier, N., & Guler, C. (1988). Beta-carotene adsorption on acid-activated mont-
enoids. Journal of Oil Palm Research, II(1), 62e78. morillonite. Journal of the American Oil Chemists Society, 65(5), 776e779. http://
Bosser, A., Paplorey, E., & Belin, J. M. (1995). A simple way to [Link]/10.1007/Bf02542530 (This paper presents a study on b-carotene
(þ/)-dihydroactinidiolide from beta-ionone related to the enzymatic coox- adsorption on acid activated montmorillonite. On this work, it was observed that
idation of beta-carotene in aqueous-solution. Biotechnology Progress, 11(6), carotenes remaining in solution after adsorption process was oxidized, giving us
689e692. [Link] the clue to our hypothesis.).
Burton, G. W., & Ingold, K. U. (1984). Beta-carotene e an unusual type of lipid Silva, S. M., Sampaio, K. A., Ceriani, R., Verhe
, R., Stevens, C., De Greyt, W., et al.
antioxidant. Science, 224(4649), 569e573. [Link] (2013). Adsorption of carotenes and phosphorus from palm oil onto acid acti-
0156. vated bleaching earth: equilibrium, kinetics and thermodynamics. Journal of
Dijkstra, A. J. (2010). Enzymatic degumming. European Journal of Lipid Science and Food Engineering, 118(4), 341e349. [Link]
Technology, 112(11), 1178e1189. [Link] 026.
Dijkstra, A., & Segers, J. (2007). Production and refining of oils and fats. In The lipid Taylor, D. R. (2005). Bleaching. In F. Shahidi (Ed.), Bailey's industrial oil & fat products
handbook with CD-ROM (pp. 143e262). CRC Press. (6th ed.), (Vol. 5); (pp. 285e340). New York: John Wiley & Son.
El-Agamey, A., & McGarvey, D. J. (2003). Evidence for a lack of reactivity of carot- Wache, Y., Bosser-DeRatuld, A., Lhuguenot, J. C., & Belin, J. M. (2003). Effect
enoid addition radicals towards oxygen: a laser flash photolysis study of the of cis/trans isomerism of beta-carotene on the ratios of volatile com-
reactions of carotenoids with acylperoxyl radicals in polar and non-polar sol- pounds produced during oxidative degradation. Journal of Agricultural
vents. Journal of the American Chemical Society, 125(11), 3330e3340. http:// and Food Chemistry, 51(7), 1984e1987. [Link]
[Link]/10.1021/Ja027986s. Jf021000g.
Fraser, M. S., & Frankl, G. (1981). Colored components of processed palm oil. Journal Zeb, A. (2012). Oxidation and formation of oxidation products of beta-carotene at
of the American Oil Chemists Society, 58(10), 926e931. [Link] boiling temperature. Chemistry and Physics of Lipids, 165(3), 277e281. http://
Bf02659660. [Link]/10.1016/[Link].2012.02.005.
1264 S.M. Silva et al. / LWT - Food Science and Technology 59 (2014) 1258e1264

Zeb, A., & Murkovic, M. (2013a). Determination of thermal oxidation and oxidation Zschau, W., & Grp, G. S. F. S. S. (2001). Bleaching of edible fats and oils e cooperative
products of beta-carotene in corn oil triacylglycerols. Food Research Interna- work of the German Society for Fat Science (DGF). European Journal of Lipid
tional, 50(2), 534e544. [Link] Science and Technology, 103(8), 505e550. [Link]
Zeb, A., & Murkovic, M. (2013b). Pro-oxidant effects of beta-carotene during thermal 9312(200108)103:8<505::Aid-Ejlt505>[Link];2e7.
oxidation of edible oils. Journal of the American Oil Chemists Society, 90(6),
881e889. [Link]