Professional Documents
Culture Documents
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem
Department of Food Technology, Faculty of Food Science & Technology, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia
Department of Food Technology, School of Industrial Technology, Universiti Sains Malaysia, 11800 Minden, Penang, Malaysia
Department of Bioprocess Technology, Faculty of Biotechnology and Biomolecular Sciences, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia
d
Sime Darby Research Sdn. Bhd., Golden Jomalina Food Industries Sdn. Bhd., Batu 9, Jalan Banting-Kelang, 42500 Teluk Panglima Garang Industrial Estate, Selangor, Malaysia
e
Department of Chemistry, College of Science, King Saud University, Riyadh 1145, Saudi Arabia
f
Institute of Biological Sciences, Faculty of Science, University of Malaya, 50603 Kuala Lumpur, Malaysia
b
c
a r t i c l e
i n f o
Article history:
Received 2 December 2011
Received in revised form 16 April 2012
Accepted 30 April 2012
Available online 11 May 2012
Keywords:
3-MCPD esters
Precursor
D-optimal design
Water degumming
Adsorbents
Crude palm oil
a b s t r a c t
The formation of 3-monochloropropane-1,2-diol (3-MCPD) esters in rened palm oil during deodorisation is attributed to the intrinsic composition of crude palm oil. Utilising D-optimal design, the effects
of the degumming and bleaching processes on the reduction in 3-MCPD ester formation in rened palm
oil from poor-quality crude palm oil were studied relative to the palm oil minor components that are
likely to be their precursors. Water degumming remarkably reduced 3-MCPD ester formation by up to
84%, from 9.79 mg/kg to 1.55 mg/kg. Bleaching with synthetic magnesium silicate caused a further 10%
reduction, to 0.487 mg/kg. The reduction in 3-MCPD ester formation could be due to the removal of
related precursors prior to the deodorisation step. The phosphorus content of bleached palm oil showed
a signicant correlation with 3-MCPD ester formation.
2012 Elsevier Ltd. All rights reserved.
1. Introduction
The presence of 3-monochloropropane-1,2-diol (3-MCPD) esters in most rened oils and fats is an important issue in the fat
and oil industry. The carcinogenicity of choropropanol shows
long-term effects in animal studies, including infertility and malfunction of certain organs, such as the kidney (Hwang, Yoon,
Kim, Jang, & Yoo, 2009). The 3-MCPD ester contents of various oils
have been reported by a number of research institutes and industry
laboratories. Rened palm oil was reported to have the highest level of 3-MCPD esters (Matthus, Pudel, Fehling, Vosmann, & Freudenstein, 2011; Zelinkov, Svejkovsk, Velek, & Dolezal, 2006).
The reason for the presence of 3-MCPD esters is unclear because
the mechanism of formation of chloroesters in a lipid environment
is not as straightforward as that in the aqueous phase.
A recent mechanism study revealed that the formation of 3MCPD esters in a hydrophobic environment, via acyloxonium ion
Abbreviations: 3-MCPD, 3-monochloropropane-1,2-diol; GCMS/MS, gas chromatography tandem mass spectrometry; CPO, crude palm oil; RBD PO, rened
bleached deodorised palm oil; DAG, diacylglycerol; MAG, monoacylglycerol; FFA,
free fatty acid; PV, peroxide value.
Corresponding author. Tel.: +60 3 89468418; fax: +60 3 89423552.
E-mail addresses: tancp@putra.upm.edu.my, tcping@yahoo.com (C.P. Tan).
0308-8146/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.foodchem.2012.04.144
800
precursors (Schurz, 2010). On the other hand, bleaching with acidactivated bleaching earth could generate the protonation of organic
compounds, such as acylglycerols, into 3-MCPD ester precursors,
imparting a strong correlation between bleaching earth acidity
and 3-MCPD ester formation during palm oil rening (Ramli et al.,
2011). However, a high dosage of activated clay can adsorb some
of the 3-MCPD ester precursors from the oil (Schurz, 2010).
In this study, the formation of 3-MCPD esters in the palm oil
rening process was assessed for a possible relationship with selected minor components of palm oil that are likely to be potential
3-MCPD ester precursors. Among the proposed precursors are partial acylglycerol (mono- and di-acylglycerol) (Rahn & Yaylayan,
2011b), phospholipids (Collier et al., 1991; Hamlet et al., 2004),
and carotenoids (ILSI, 2009). Use of a D-optimal design modelling
of the experiment (involving a categorical factor) is possible with
a reduced number of experimental sets (Hoskins, Colbourn, &
Montgomery, 2005). This method allows a detailed analysis of
the effect of related 3-MCPD ester precursors on the level of 3MCPD esters formed in the nal rened oil when different degumming and bleaching treatments are employed.
2. Materials and methods
2.1. Materials
Crude palm oil (CPO) was supplied by Sime Darby Golden Jomalina Sdn. Bhd. (Teluk Panglima Garang, Selangor). The chemical,
3-chloropropane-1,2-diol (3-MCPD), was purchased from Aldrich
(France), and isotopically labelled 3-chloropropane-1,2-diol (3MCPD-d5) internal standard was purchased from Cambridge Isotope Laboratories (CIL, Massachusetts, USA). Phenylboronic acid
(PBA) of 99.0% purity was purchased from Fluka (Shanghai, China). All other chemicals and solvents were of analytical grade.
Synthetic magnesium silicate, Magnesol R60 (Dallas Group of
America, Inc., New Jersey, USA), was provided by Global Speciality
Ingredients (M) Sdn. Bhd. (Port Klang, Selangor) as a Malaysian
representative of Dallas Group of America, Inc. (New Jersey,
USA). Activated carbon was provided by Efgen Carbon Sdn Bhd
(Kapar, Selangor). Taiko Supreme 1B activated clay (Taiko Group
Sdn. Bhd, Pasir Gudang, Johor), Taiko TPT 110 non-activated natural clay (Taiko Group Sdn. Bhd, Pasir Gudang, Johor) and phosphoric acid, 85% (Merck, Darmstadt, Germany), were supplied
by Sime Darby Golden Jomalina Sdn. Bhd. (Teluk Panglima
Garang, Selangor).
2.2. Laboratory-scale physical rening
Four different grades of CPO were rened with standard conditions that are commonly practised in the industry. Approximately
700 g of crude palm oil were degummed and bleached in a 2 l lab-
Table 1
Quality characteristics and composition of four different crude palm oils.
Quality characteristics
Standard quality I
Standard quality II
<LOD
1.54 ( 0.08)a
1.19 ( 0.02)a
Nil
3.2 ( 0.04)b
3.8 ( 0.3)a
615 ( 1.0)b
3.72 ( 0.02)b
0.0a
<LOD
1.49 ( 0.05)a
1.34 ( 0.02)a
Nil
3.0 ( 0.04)b
4.4 ( 0.4)a
611 ( 1.0)b
3.65 ( 0.02)b
0.0a
<LOD
1.72 ( 0.22)a
3.52 ( 0.03)b
Nil
2.9 ( 0.08)b
4.8 ( 0.3)a
597 ( 1.4)b
5.63 ( 0.02)c
0.05 ( 0.02)a
0.061 0.01
5.93 ( 0.41)b
4.19 ( 0.01)b
0.90 0.01
2.4 ( 0.12)a
8.8 ( 0.3)b
476 ( 1.2)a
2.20 ( 0.02)a
0.54 ( 0.01)b
801
Independent variablesa
Percent acid
(X1, %)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
a
0
0
0
0
0
0.05
0.05
0.05
0.05
0.1
0.1
0.1
0.1
0.1
0.1
Bleaching
adsorbent (X2)
A
B
C
D
D
A
B
C
D
A
A
B
C
C
D
Response variables
Characteristic of RBD palm oil
3-MCPD
(Y1, mg/kg)
Colour
(Y2, R)
DAG
(Y3, %)
MAG
(Y4, %)
Carotene
(Y5, ppm)
Phosphorus
(Y6, ppm)
1.55 0.05
1.75 0.08
4.32 0.08
0.46 0.05
0.51 0.07
9.79 0.05
10.3 0.05
14.0 0.06
8.24 0.04
9.28 0.07
10.5 0.1
10.1 0.04
11.7 0.1
9.75 0.04
7.50 0.05
4.0
4.4
5.9
4.2
4.3
3.1
3.4
5.8
3.2
2.0
2.0
2.5
5.2
5.2
2.3
3.38 0.02
2.55 0.03
2.21 0.04
2.50 0.03
2.45 0.02
2.95 0.02
2.45 0.03
2.26 0.02
2.45 0.03
2.36 0.01
2.35 0.01
2.21 0.02
2.32 0.02
2.35 0.02
2.68 0.03
0.35 0.02
0.18 0.01
0.16 0.02
0.14 0.01
0.12 0.01
0.31 0.01
0.25 0.01
0.19 0.01
0.15 0.01
0.42 0.01
0.39 0.01
0.19 0.02
0.18 0.02
0.20 0.02
0.18 0.01
436 1.0
408 0.9
425 1.0
454 2.0
461 2.0
389 2.6
391 2.0
405 2.0
443 2.7
375 2.1
382 1.9
380 2.5
405 1.4
408 1.9
438 1.5
4.52 0.08
4.66 0.10
7.77 0.09
3.01 0.09
3.02 0.09
5.98 0.08
8.26 0.05
16.1 0.14
6.78 0.10
7.77 0.12
7.70 0.10
10.7 0.08
20.8 0.09
22.0 0.10
6.51 0.07
For factor X2, A, activated clay; B, natural clay; C, activated carbon; D, magnesium silicate.
Y i b0 bi X i bii X 2i bij X i X j ;
where Yi is the response and b0, bi, bii, and bij are the constant coefcients of intercept, linear, quadratic and interaction terms, respectively. Xi and Xj are independent variables. Each model was analysed
using ANOVA, a regression analysis and graphical plots. The adequacy of the models was evaluated by the lack of t test, F-test
and ANOVA, and the overall predictive capability of the model
was explained by the coefcient of determination R2.
All of the experimental results in this study are reported as the
means standard deviations (SD) of triplicate analyses. Signicance was established at p < 0.05 for Tukeys test for ANOVA, using
MINITAB statistical software (Version 14, Minitab Inc., PA, USA).
The Pearson correlation coefcients between different response
variables and the 3-MCPD content were analysed using the same
statistical software programme.
2.4. 3-MCPD ester determination using GCMS/MS
The relative 3-MCPD ester contents of CPO and RBD palm oil of
each experimental run were determined using the modied BfR
Method 008 (BfR. Federal Institute for Risk Assessment: BfR Method_82_FC-008-01, 2009). The determination of 3-MCPD esters was
carried out based on an indirect method using acid transesterication, which involves the release of 3-MCPD from its esters, purication by ammonium sulphate extraction, PBA derivatisation and
quantication using gas chromatography with tandem mass spectrometry detection (GCMS/MS). A quantitative analysis of the derivatised 3-MCPD compound on a Thermo TSQ Quantum XLS Triple
Quadruple GCMS/MS system (Waltham, MA, USA) was carried out
by monitoring characteristic ions (quantier) at m/z 147 of 3MCPD and m/z 150 of isotopically labelled internal standard (3MCPD-d5). Ions at m/z 91 and 196 (3-MCPD) and m/z 93 and 201
(3-MCPD-d5) were used as qualiers. The 3-MCPD content was calculated as the ratio of the peak area responses of 3-MCPD (m/z
147) over the internal standard (m/z 150) for the calibration standards, blank and samples. The respective content was determined
from a matrix-based calibration graph that was constructed by
plotting the peak area ratios for the calibration standards (y-axis)
against the amount of 3-MCPD (x-axis).
Further analysis was carried out to detect the effect of increasing deodorisation temperature on 3-MCPD ester formation in Standard Quality II CPO. As shown in Fig. 1, the formation of 3-MCPD
esters signicantly increased as the deodorisation temperature increased. The level of chloroesters reached 9 mg/kg when the sample was deodorised at 250 C. Further increases in temperature,
however, led to the degradation of 3-MCPD esters, as shown previously (Matthus et al., 2011; Pudel et al., 2011).
3.3. Effects of degumming and bleaching on the reduction of 3-MCPD
esters
3.3.1. General
The degumming and bleaching processes are steps that could
potentially reduce 3-MCPD ester formation by removing related
precursors prior to deodorisation. Utilising D-optimal design, the
effects of different acid dosages and bleaching adsorbents on the
reduction of 3-MCPD esters were scrutinised, using low-grade
CPO (Standard Quality II). The deodorisation temperature was set
at 250 C because the maximum formation of 3-MCPD esters occurred at this temperature.
3.3.2. Model tting
Table 2 shows the responses of D-optimal design, namely 3MCPD content (Y1), the colour (Y2) of RBD palm oil, and the BPO
DAG (Y3), MAG (Y4), carotene (Y5) and phosphorus (Y6) contents.
All responses were signicantly tted into reduced quadratic polynomial models (p < 0.05). The best tting models were determined
by a regression analysis with backward elimination in which insignicant factors were deleted from the initial model to yield the
most signicant nal reduced models. The ANOVA results of the nal reduced models, the p-value of signicant terms, the lack of t
test, the F-test and the determination coefcient R2 of the responses are shown in Table 3.
The 3-MCPD levels (Y1) were signicantly (p < 0.05) modelled
into a polynomial model with signicant quadratic terms. The lack
of t of the model was not signicant (p > 0.05). The R2 value was
0.974 with no observed outliers, suggesting that the data were well
tted to the described model. The computed model F-value of 27.9,
in mg/kg of 3-MCPD, was higher than the tabular value of
F0.05(6,3) = 8.94, implying that the model was signicant at the
95% condence level. For the other responses, the F-values of the
nal reduced model ranged from 35.46 to 255.23, and the R2 values
were close to one, implying that the models are signicant.
3.4. Effects of degumming and bleaching on the formation of 3-MCPD
esters
The formation of 3-MCPD esters during deodorisation was inuenced by the degumming and bleaching parameters, as depicted in
Fig. 2(a). Water degumming signicantly reduced 3-MCPD ester
3-MCPD esters (mg/kg)
802
12
d
10
6
b
a
4
2
0
180 C
200 C
230 C
250 C
270 C
Temperature (C)
Fig. 1. The 3-MCPD esters level of rened bleached deodorised palm oil (RBD PO) at
different deodorisation temperatures. (Same letters do not differ signicantly at
p 6 0.05).
803
Model
p-value
F-value
X1
X2
X1X2
X12
R2
Adjusted R2
Predicted R2
Lack of t (p-value)
Lack of t (F-value)
Quadratic
0.0002a
36.45
<0.0001a
0.0081a
0.481
0.0001a
0.980
0.953
0.829
0.482
1.10
2FI
<0.0003a
27.72
<0.0001a
<0.0001a
0.0003a
0.997
0.995
0.982
0.053
8.77
2FI
<0.0001a
60.42
0.0001a
<0.0001a
<0.0001a
0.984
0.967
0.873
0.054
8.62
Linear
<0.0001a
35.46
0.0481a
<0.0001a
0.934
0.908
0.855
0.152
3.75
Quadratic
<0.0001a
42.57
0.0003a
<0.0001a
0.0093a
0.0182a
0.983
0.960
0.789
0.244
2.42
Quadratic
<0.0001a
138.36
<0.0001a
<0.0001a
0.0004a
0.0460a
0.995
0.987
0.942
0.207
2.85
DESIGN-EXPERT Plot
Interaction graph
B: Absorbents
14.0
DESIGN-EXPERT Plot
X = A: Acid %
Y = B: Absorbents
Design Points
B1 Activated Clay
B2 Natural Clay
B3 Activ Carbon
B4 Mg silicate
6.5
393
-0.9
369
0.03
(a)
0.05
0.08
0.10
Interaction graph
0.05
0.08
0.10
Interaction graph
Color (R)
X = A: Acid %
Y = B: Absorbents
B: Absorbents
6.1
X = A: Acid %
Y = B: Absorbents
Phosphorus (ppm)
Design Points
17.5
Design Points
12.4
B1 Activated Clay
B2 Natural Clay
B3 Activ Carbon
B4 Mg silicate
4.0
2.9
2.3
1.9
0.03
0.05
A: Acid %
0.08
5.1
7.3
0.00
(b)
0.03
A: Acid %
DESIGN-EXPERT Plot
B: Absorbents
22.5
0.00
(c)
A: Acid %
DESIGN-EXPERT Plot
B1 Activated Clay
B2 Natural Clay
B3 Activ Carbon
B4 Mg silicate
417
2.8
0.00
Phosphorus (ppm)
441
Carotene (ppm)
10.3
Color (R)
B1 Activated Clay
B2 Natural Clay
B3 Activ Carbon
B4 Mg silicate
B: Absorbents
465
X = A: Acid %
Y = B: Absorbents
Design Points
Interaction graph
Carotene
0.00
0.10
(d)
0.03
0.05
0.08
0.10
A: Acid %
Fig. 2. Interaction plots showing the effects of degumming and bleaching treatments on: (a) 3-MCPD levels in rened bleached deodorised palm oil (RBD PO), (b) phosphorus
content and, (c) carotene content in bleached palm oil (BPO), and (d) colour of RBD PO.
DESIGN-EXPERT Plot
Correlation: -0.393
10.6
7.2
3.9
0.5
2.21
(a)
2.50
2.79
3.09
3.38
DESIGN-EXPERT Plot
Correlation: 0.683
14.0
formation from 9.79 to 1.55 mg/kg (p < 0.05). This reduction could
be because reactants such as chloride, which contributes to the formation of 3-MCPD esters, are washed off during water degumming
(Matthus et al., 2011). The phosphoric acid dosage had a greater
effect on the development of chloroesters than did the different
types of bleaching adsorbents. The increase in acid dosage had
no signicant effect on the 3-MCPD ester level (p > 0.05), with
the exceptions of the activated carbon and magnesium silicate
conditions.
Bleaching, using magnesium silicate (Magnesol R60), showed
the greatest reduction of 3-MCPD esters compared to activated
clay, natural clay and activated carbon, especially when water
degumming was used. Magnesium silicate possesses a relatively
large surface area and the most active basic sites compared to alumina, silica and activated carbon (Zhu, Yates, & Caldwell, 1994).
These properties may be responsible for the adsorption or degradation of particular precursors of 3-MCPD esters from BPO prior to
deodorisation.
Haines et al., (2011) indicated that 3-MCPD was unstable above
pH 6. High pH conditions may assist in the removal of related precursors and cause a reduction of 3-MCPD esters. However, when
acid was used during degumming, the pH conditions of bleaching
dropped and reduced the removal effect. The pH values of the
bleaching adsorbents were tested at 6.0, 9.0, 9.5 and 9.4 for activated clay, natural clay, activated carbon and magnesium silicate,
respectively. However, the effect of pH on 3-MCPD ester reduction,
as reported previously (Ramli et al., 2011), was only signicant for
magnesium silicate. Bleaching with natural clay slightly increased
3-MCPD ester formation compared to activated clay, which possessed a much lower pH. This nding may also suggest that acid
activation of bleaching earth, using sulphuric acid, does not contribute to the activation of 3-MCPD ester precursors during bleaching. On the other hand, bleaching with activated carbon resulted in
the formation of high amounts of MCPD esters, up to 14 mg/kg. A
similar observation was reported by Haines et al., (2011) but with
higher amounts of 3-MCPD esters because a direct method of 3MCPD ester detection was used. Although activated carbon possesses the highest pore volume, which allows greater adsorption
of precursors and impurities, the chemistry of the precursors and
their mechanisms of adsorption could be among the factors that
require further investigation.
The quality characteristics of RBD palm oils were tested and
compared. PV was not present for all rening treatments, indicating that all of the rening treatments sufciently remove oxidative
compounds from crude oil, and no signicant difference was observed for the FFA content (p > 0.05). Water degumming produced
a strong colour, in the range of 4.0 to 5.9 red, suggesting that colour
xation might occur during deodorisation due to an insufcient
degumming treatment. Activated clay resulted in excellent pigment removal, while magnesium silicate caused a slightly stronger
colour, as depicted in Fig. 2(d).
14.0
804
10.6
7.2
3.9
0.5
3.0
(b)
7.8
12.5
17.2
22.0
Phosphorus (ppm)
Fig. 3. Correlation of 3-MCPD levels with (a) diacylglycerol, and (b) phosphorus
content in bleached palm oil (BPO).
805