EFFECTS OF EMULSIFIERS ON CRYSTAL BEHAVIOR OF PALM

OIL BLENDS ON SLOW CRYSTALLIZATION

M.S. MISKANDAR1,2, Y.B. CHE MAN1,4, R. ABDUL RAHMAN1, I. NOR AINI2

and M.S.A. YUSOFF3
1
Department of Food Technology
Faculty of Food Science and Technology
Universiti Putra Malaysia
43400 UPM, Serdang, Selangor, D.E., Malaysia
2
Malaysian Palm Oil Board
No. 6 Persiaran Institusi
Bandar Baru Bangi
Kajang, Selangor, D. E., Malaysia
3
Golden Hope Research
Banting Selangor, Malaysia

Submitted for Publication April 13, 2006

Revised Received and Accepted July 20, 2006

ABSTRACT

The effects of sorbitan tristearate (STS) and soy lecithin on blends of

palm oil (PO) and olein by slow crystallization were studied. The blends of PO
and olein were melted at 80C for 30 min and slowly cooled at a rate of
0.7C/min for 32 min and at 20 ⫾ 0.1C for 48 min. Nuclear magnetic reso-
nance was used to measure the solid fat content (SFC) development. Crystal-
lization without emulsifier was influenced by trisaturated and disaturated
triacylglycerol content of PO manifesting significant (P < 0.05) changes in
SFC, crystal dispersion and viscosity, but emulsifiers in the blends disturbed
these behaviors. The reaction of lecithin was very selective and was deter-
mined that at 0.03%, it generally served as a crystal promoter, while at
0.06–0.09%, it was a crystal inhibitor. Viscosity was closely related to the
crystal dispersion in the liquid. The blends had low viscosity with 0.03%
lecithin, while blends with high trisaturate content had high viscosity with
0.06% lecithin. STS was generally a crystal inhibitor, although at various
levels, it was a crystal promoter for the middle-range saturates. Slow crystal-
lization process of PO and olein blends with 0.03% lecithin produced homog-
enous crystals with low SFC and viscosity for fluidized products.
4
Corresponding author. TEL: 603-89468413; FAX: 603-89423552; EMAIL: yaakub@
fsb.upm.edu.my

Journal of Food Lipids 14 (2007) 1–18. All Rights Reserved.

© 2007, The Author(s) 1
Journal compilation © 2007, Blackwell Publishing
2 M.S. MISKANDAR ET AL.

INTRODUCTION

The physical state at which hard fat crystallizes is important (O’Brien

2000). However, it is not an ultimate effect of the current crystallization
conditions alone, but depends on the entire history of the crystallization
process (Blaurock 1999; Jahaniaval et al. 2000; Miskandar et al. 2002). Hot oil
that is rapidly cooled to a temperature below a melting point would most
probably crystallize in a form followed by the formation of b′ and b crystals
(Blaurock 1999). However, on slow cooling, b′ crystal would first form fol-
lowed by the a crystal surrounding the b′ crystal that acts as nuclei for
crystallization (Blaurock 1999). Besides various temperature conditions,
emulsifier with different hydrophobic properties would also affect fat proper-
ties by accelerating or retarding crystallization and polymorphic transforma-
tion (Garti 2002; Miskandar et al. 2002).
Emulsifiers are important ingredients in bakery shortening, although only
a small amount is used, and typically at <0.5% can contribute to significant
product qualities (Stauffer 1996). According to Stauffer (1996), Nash and
Lawrence (1972) and Chrysam (1985), emulsifiers in food products are mul-
tifunctional and contribute to the final product properties. For the purpose of
certain functionalities, fluid shortening can act as the carrier for these emul-
sifiers and later becomes an ingredient for the production of food products
such as bread and cake, dough strengtheners and antistaling agent (O’Brien
2000).
Emulsifiers are associated with triacylglycerol (TAG) by their hydropho-
bic groups (fatty acid moieties), specifically the acyl–acyl interaction (Kat-
suragi 1999). The acyl group of emulsifiers determines its functionality with
respect to the TAG. The major effects of emulsifiers on fat crystal occur in the
area of fat nucleation, transformation of polymorphism and crystal growth.
These effects alter the physical properties of fat such as crystal size, solid fat
content (SFC) and crystal arrangements. The question whether an emulsifier
will promote or retard crystallization in oils and fats is very subjective and
depends on various factors. However, Katsuragi (1999) suggested that similar
acyl groups promote crystallization, while different acyl groups retard crystal
development. Seeding effect could be another phenomenon on the effect of
emulsifiers in fluid shortening by promoting crystallization (O’Brien 2000).
Sorbitan tristearate (STS) is a hydrophobic emulsifier with a hydrophobic
lipophilic balance (HLB) value of 2.1 (Kamel 1997). The molecular structure
of STS is very similar to that of TAG, thus enabling STS to cocrystallize with
fats and form solid solutions (Krog and Last 1995). Lecithin, on the other
hand, is a complex phospholipid that has hydrophilic and hydrophobic prop-
erties, with an HLB of 3.5 (O’Brien 2000). The main function of phosphatide
is to emulsify fats. The long-chain fatty acid moieties that contribute the
 EFFECTS OF EMULSIFIERS ON PO 3

hydrophobic property are counterbalanced by the polar or hydrophilic prop-

erties of the phosphate moiety. Emulsifiers can also act as wetting agents in
fluid shortening systems by reducing the interfacial tension between the liquid
and solid surfaces that cause the liquid to spread more evenly over surfaces. As
agglomeration of crystal is not desired for a stable fluid shortening, emulsifiers
can be used to control the fat particle coagulation (O’Brien 2000).
The aim of this study was to determine the crystallization properties of 11
different blends of palm oil (PO) and palm olein using STS and lecithin.

MATERIALS AND METHODS

Materials
Refined, bleached and deodorized (RBD) PO and palm olein were
obtained from the Malaysian Palm Oil Board, Malaysia. STS was from
Danisco Malaysia Sdn. Bhd., Penang, Malaysia, and lecithin was bought from
Euro Chemopharma Sdn. Bhd., Shah Alam, Selangor, Malaysia.

Physical Analyses
RBD PO and palm olein were melted at 80C in a Memmert drying oven
(model 854 UL 80; Memmert GmbH, Schwabach, Germany). STS and leci-
thin were dissolved in PO before mixing in the blends to assist homogeneity.
Eleven blends of PO and palm olein were prepared in three replications in
the ratios of 100:0, 90:10, 80:20, 70:30, 60:40, 50:50, 40:60, 30:70, 20:80,
10:90 and 0:100 (w/w), and these were represented by the letters A, B, C, D,
E, F, G, H, J, K and L, where A and L were 100% PO and palm olein,
respectively. The experimental preparation was for the study of the effects of
STS and lecithin at 0, 0.03, 0.06 and 0.09% by slow crystallization. Each
sample was filled in nuclear magnetic resonance (NMR) tube of 8-mm diam-
eter to a height of 30 mm for SFC measurement and tall-formed beaker
(69-mm internal diameter, 130-mm height) to a height of 45 mm for viscos-
ity tests. The tubes and beakers were placed in the water bath at 80C for
30 min to destroy all the crystal memory (De Man et al. 1989). SFC mea-
surement took an average of 2.5 s, while viscosity measurement needed 35 s
per sample (included in the time was sample preparation and measurement).
The rest of the tubes and beakers were immediately transferred into the
cooling bath at crystallization temperature (Tc) of 20C for crystallization
time (Tm) of 80 min (cycle 1). The average time for the temperature to drop
from 80 to 20C was 32.08 ⫾ 2.0 min (0.70C/min).
The SFC at the end of the study was determined by pulse NMR (model
NMS120, Minispec Bruker, Rheinstetten, Silberstreifen, Germany, with refer-
4 M.S. MISKANDAR ET AL.

ence samples of triolein and tristearin also in NMR tubes). Previous studies
had shown that the crystallization Tc 20C for Tm 80 min was sufficient for full
crystallization of PO: olein blends with 50% and more PO.
The viscosity test was carried out by a Brookfield digital viscometer
(model DV11, Brookfield Engineering Laboratories, Inc., Stoughton, MA)
using a speed of 50 rpm for 25 s. The viscosity of the samples in centipoises
was taken at the end of each tempering session.
Microscopic examination for crystal distribution was measured by
placing a drop of the tempered samples, from the same sample set previously
mentioned, on a glass slide. A glass cover slip was placed on top of the sample,
and the sample was immediately viewed under a polarized microscope
(Olympus model BH-2, Olympus Corporation, Tokyo, Japan) attached to a
video color camera. Leica (Leica Q500mc Qwin vol 0.02; Leica Cambridge
Ltd, Cambridge, U.K.) for windows was used to analyze the images.
TAG composition was determined by high-performance liquid chroma-
tography according to the American Oil Chemist Society (AOCS 1989)
method. The mobile phase used was acetone/acetonitrile (75:25, v/v). The
sample was diluted in acetone to make up to 10% sample concentration. A
volume of 10 mL was injected into a Waters C18 column (3.9 ¥ 300.0 mm).
The oven temperature was set at 35C, and the flow rate of the solvent was
1.5 mL/min. All sample determinations were carried out in duplicate.

Statistical Analysis
Analysis of variance at 95% confidence level and multiple comparisons
by Duncan was determined in statistical software package SPSS for Windows
version 11.5. Arrangement of data for the statistical analysis was done in
Microsoft Excel 2000 (Microsoft Corporation, Redmond, WA).

RESULTS AND DISCUSSION

Crystallization of Fat Blend without Emulsifiers

When blends of PO and olein were slowly cooled from 80 to 20C, they
underwent an average temperature change of 0.7C/min for the first 33 min and
20.0 ⫾ 0.1C for the rest of 47 min (cycle 1), and created the condition of slow
crystallization, even to blends with high trisaturate and disaturate TAGs (Mis-
kandar et al. 2004). No significant change on the temperature of the blend and
water bath was recorded during the crystallization process. The change in the
temperature was so slow, which caused slow crystallization, and any tempera-
ture changes due to crystallization were not significant.
 EFFECTS OF EMULSIFIERS ON PO 5

NMR was used to determine the crystallization properties of blends A–L

in this study owing to its ability for direct SFC measurement. Blaurock (1999),
Busfield and Proshogo (1990) and Sugimoto et al. (2001) had shown that
NMR provides an important tool for measuring fat crystallization. Albeit
studies on fat crystallization had used X-ray diffraction and differential scan-
ning calorimeter in trilaurin and tripalmitin crystallization, X-ray diffraction
could not be used under the present experimental conditions because of the
fluid nature of most of the samples.
Controlling the Tc is an important procedure in crystal development. The
PO and olein blends in this study were tempered at Tc of 80C for Tm 30 min to
provide an environment that was free of crystal memory and with the same
starting history of 0% SFC. According to Blaurock (1999), crystallization
history of the fat is considered an important step in fat product preparation, as
it contributes to the final crystal type.
As the fat underwent cooling treatment for Tm 80 min, crystallization
started to occur when the supercooling temperature was low enough for the
highest melting crystals to create the nuclei for the centers of crystallization
(Faur 1996; Ghotra et al. 2002). The greater the number of nuclei formed, the
more rapid the whole bulk will crystallize. According to Blaurock (1999), in an
instantaneous cooling of melted oil to a temperature below the melting of a
type of crystal, the crystallization will be of that type of crystal. The cooling
rate of 0.7C/min was a slow cooling process that initiated the high melting
crystal types to crystallize first, and because the cooling time was sufficiently
long, big crystal aggregates were formed.
Slow crystallization at 0.7C/min for cooling time Tm 80 min in an earlier
study had provided sufficient crystallization time for blends with high SSS
(Miskandar et al. 2004). The condition was suitable for PO and olein blends
with no emulsifiers. However, the presence of emulsifiers in these blends had
influenced their crystallization behaviors, as shown by change of SFC
(Table 1). The SFC of blends of PO, from 0 (Blend L) to 100% (Blend A)
ranged from 1.52 to 8.92%, in a nonlinear fashion as shown by the equation
of SFC = -0.0202X3 + 0.03476X2 - 0.5313X% PO and R2 = 0.9846. As the
content of PO was reduced to lower than 40%, the changes in SFC were not
significant (P > 0.05) because the blends with and without emulsifiers at that
level were practically in the liquid form. Crystallization behavior and solid fat
development in blends with more than 40% PO, however, were dependent on
the PO content, in agreement with findings of Martini et al. (2002) on crys-
tallization with high melting TAGs.
Crystal Distribution without Emulsifiers
The formation of crystals at the end of cycle 1 showed that vigorous
crystal development had occurred in blends without emulsifiers as shown
 6

TABLE 1.
THE EFFECTS OF EMULSIFIERS ON THE SOLID FAT CONTENT OF PALM OIL AND OLEIN BLENDS BY CYCLE 1

Blend Different emulsifiers

0% Lecithin (% w/w) Sorbitan tristearate (% w/w)

0.03 0.06 0.09 0.03 0.06 0.09

A 9.5 ⫾ 0.24Cj 9.7 ⫾ 0.14CDg 9.1 ⫾ 0.10Bg 8.8 ⫾ 0.07Abi 9.8 ⫾ 0.14CDg 9.4 ⫾ 0.04BCg 9.4 ⫾ 0.16BCh
B 9.2 ⫾ 0.23Di 9.1 ⫾ 0.10DCf 8.5 ⫾ 0.07Bfg 8.3 ⫾ 0.13Abh 9.2 ⫾ 0.12CDf 8.8 ⫾ 0.22BCf 8.4 ⫾ 0.25Abg
C 8.0 ⫾0.22Abh 8.4 ⫾ 0.15Bef 7.8 ⫾ 0.06Aef 7.9 ⫾ 0.02Ag 8.1 ⫾ 0.23Abe 8.0 ⫾ 0.23Abe 7.7 ⫾ 0.13Af
D 7.6 ⫾ 0.05BCg 7.9 ⫾ 0.30Def 7.6 ⫾ 0.15BCe 7.3 ⫾ 0.10Abf 7.8 ⫾ 0.14Cde 7.5 ⫾ 0.09Bd 7.0 ⫾ 0.19Ae
E 6.4 ⫾ 0.57Abf 6.9 ⫾ 0.50Cd 6.1 ⫾ 0.44Ad 6.7 ⫾ 0.02Bce 6.7 ⫾ 0.69BCd 6.4 ⫾ 0.73Abc 6.3 ⫾ 0.46Abd
F 5.5 ⫾ 0.36Bbe 6.7 ⫾ 0.25Dd 5.4 ⫾ 0.25ABd 6.3 ⫾ 0.09CDd 6.4 ⫾ 0.08CDd 5.9 ⫾ 0.23Cc 5.2 ⫾ 0.12Abc
G 4.2 ⫾ 0.09BCd 5.7 ⫾ 0.35Dc 3.5 ⫾ 0.40Ac 5.5 ⫾ 0.10Dc 4.7 ⫾ 0.35CDc 4.2 ⫾ 0.10BCb 4.3 ⫾ 0.07BCb
H 1.6 ⫾ 0.15Bc 2.6 ⫾ 0.74DEb 2.1 ⫾ 0.12CDb 2.5 ⫾ 0.21Deb 2.6 ⫾ 0.27DEb 0.1 ⫾ 0.00Aa 0.1 ⫾ 0.01Aa
M.S. MISKANDAR ET AL.

J 0.3 ⫾ 0.02Ab 0.1 ⫾ 0.02Aa 0.1 ⫾ 0.01Aa 0.1 ⫾ 0.03Aa 0.1 ⫾ 0.02Aa 0.1 ⫾ 0.01Aa 0.1 ⫾ 0.05Aa
K 0.3 ⫾ 0.12Ab 0.1 ⫾ 0.01Aa 0.1 ⫾ 0.06Aa 0.1 ⫾ 0.01Aa 0.1 ⫾ 0.02Aa 0.1 ⫾ 0.01Aa 0.1 ⫾ 0.07Aa
L 0.1 ⫾ 0.08Aa 0.1 ⫾ 0.02Aa 0.1 ⫾ 0.02Aa 0.1 ⫾ 0.01Aa 0.1 ⫾ 0.02Aa 0.1 ⫾ 0.02Aa 0.1 ⫾ 0.01Aa

Mean of three determinations ⫾ SD.

A–E
Mean with different letters within each row is significantly different (P < 0.05).
a–j
Mean with different letters within each column is significantly different (P < 0.05).
Cycle 1, treatment temperature and time of 80C (30 min) → 20C (80 min).
 EFFECTS OF EMULSIFIERS ON PO 7

50 μm
A B
FIG. 1. PHOTOMICROGRAPH OF CRYSTAL DISPERSION OF PALM OIL AND OLEIN
BLEND ON SLOW CRYSTALLIZATION (10 ¥ 10 MAGNIFICATION)
A and B are Blends B and E, respectively, without emulsifier.

under a polarized light microscope (PLM) in Fig. 1A,B. Complete crystalli-

zation had occurred in Blend B at Tm 80 min, as shown by the formation of
mixtures of crystal sizes and shapes covering the whole range of the slide
(Fig. 1A). The study showed that crystallization not assisted by stirring had
created the conditions for crystallization that favored the nearest nuclei. The
nuclei could be any foreign matter, the high melting TAGs or other already
formed crystals.
Blend E containing high oleic acid was a 1:1 mixture of PO and olein. As
shown in Fig. 1B, fine crystals were distributed more evenly in this blend than
in Blend B. The liquid TAGs (triunsaturated TAG and diunsaturated TAG)
were in bigger quantity in Blend E, thus influencing crystals to move more
freely and creating more crystallization sites that improved the rate of crys-
tallization. The condition had caused crystals to be finely formed and homo-
geneously distributed throughout the blend. As the ratio of the liquid TAGs in
Blends G, H, I, J and K was increased, crystallization was not apparent under
PLM. It showed that in cycle 1, Tc and Tm were insufficient to start crystalli-
zation and crystal development. Thus, Tc should be reduced and Tm increased.

SFC, Viscosity and Crystal Behavior without Emulsifier

The solid fat development of Blends B and E followed the same trend as
their viscosities and crystal dispersions in the absence of an emulsifier. Crystal
formations were greatly affected by the content of PO in the blend. As PO was
reduced from 100 to 0% in Blends A–L, the SFC and viscosity were reduced
8 M.S. MISKANDAR ET AL.

10 1200

8 1000
SFC
VISC
6 800
R2 = 0.9846

Viscosity (cps)
SFC (%)

4 600

2 400
R2 = 0.9689

0 200
A B C D E F G H I J K

-2 0
PO Blend with 0% emulsifier

FIG. 2. SOLID FAT CONTENT (SFC) (%) AND VISCOSITY (VISC, CPS) AS A FUNCTION OF
PALM OIL (PO) AND PALM OLEIN BLENDS AT SLOW CRYSTALLIZATION

almost linearly as shown in Fig. 2. The high PO content had created viscosity
with high centipoise in Table 2 as well as densely packed crystals as shown in
Fig. 1A. In the same respect, the crystal dispersion became loosely packed as
the proportion of olein was increased in Blend E (Fig. 1B), and this was
followed by a significant drop in viscosity from 832 to 485 cps.

Effect of Lecithin on Crystal Behavior

The optimum amount of STS and lecithin effectively retarding or pro-
moting crystal growth of Blends A–L was influenced by Tc and Tm. In general,
at 0.03% level, lecithin had caused SFC to increase, but at 0.06 and 0.09%,
lecithin caused SFC to decrease. As shown in Table 1, lecithin had caused an
increase (P < 0.05) in the SFCs of Blends C–H, but no significant change was
observed in the SFCs of A, B, J, K and L at 0.03%. Increasing the lecithin level
to 0.06% caused a reduction (P < 0.05) in the SFCs of Blends A, B, G and H,
but no significant change in other blends was observed. At 0.09% lecithin,
SFCs were reduced in Blends A–H, while other blends remained unchanged.
Table 1 shows that there was possible preference in the interaction of lecithin
at its phosphate ends with the acyl ends of TAGs (Katsuragi 1999). It was
suggested that at 0.03% lecithin, the phosphate ends had interacted with a low
 EFFECTS OF EMULSIFIERS ON PO 9

TABLE 2.
THE EFFECTS OF LECITHIN ON THE VISCOSITY (CPS) OF
PALMOIL AND OLEIN BLENDS BY CYCLE 1

Blend % Lecithin

0 0.03 0.06 0.09

B 832 ⫾ 55Cc 394 ⫾ 64Ab 725 ⫾ 56Bc 1343 ⫾ 28Db

E 485 ⫾ 102Bb 112 ⫾ 0Aa 384 ⫾ 27Bb 186 ⫾ 24Aa
G 138 ⫾ 9Ca 101 ⫾ 9Aa 117 ⫾ 9Ba 112 ⫾ 0ABa

Mean of three determinations ⫾ SD.

A–D
Mean with different letters within each row is significantly dif-
ferent (P < 0.05).
a–c
Mean with different letters within each column is significantly
different (P < 0.05).
Cycle 1, treatment temperatures and time of 80C (30 min) → 20C
(80 min).

number of acyl groups of the TAGs that retarded the crystallization and growth
of the infected surface and crystal, but allowed the uninfected surface and
crystals to grow as the blend was cooled. Crystal growth and crystal size was
enhanced, as the infected crystals became the centers of crystallization. In
general, it was noted that the SFC increase was only in blends containing
0.64–1.03% trisaturates (middle range of trisaturates), while no significant
change was noted in blends with a high range of trisaturates. At 0.06 and
0.09% lecithin, increasing points of interaction had occurred on more acyl
ends, allowing less crystallization to take place. Retardation of crystal growth
had occurred in most blends as shown by the reducing SFCs as lecithin was
increased from 0.06 to 0.09% in the experiments as shown in Table 1.
The significant change in crystal development, as shown by the SFCs,
was well correlated with crystal distribution. It was densely formed in Blend
B without emulsifier, but the addition of lecithin had reduced the crystal size
(Fig. 3A–C) due to physical and chemical interactions. The presence of
0.09% lecithin had induced the formation of the most crystal in comparison
to the rest of the combinations in Blend B. As more liquid was present in
Blend E, crystal arrangement became loosely packed in relation to the reduc-
ing amount of trisaturates in the blends. Lecithin had greatly reduced
the crystal size of the blends; however, the amount of lecithin added into the
blend had increased the crystal growth as shown by the crowdedness of the
crystal under PLM (Fig. 3D–F).
Zaki et al. (1997) determined that tripalmitoyl-glycerol (PPP), sn-1-
myristoyl-2,3-dipalmitoyl-glycerol (MPP) and sn-1,2-dipalmitoyl-3-stearoyl-
glycerol (PPS) are responsible for influencing crystallization in palm
10 M.S. MISKANDAR ET AL.

A D

B E

C F
FIG. 3. PHOTOMICROGRAPH OF EFFECT OF LECITHIN ON CRYSTAL DISPERSION OF
PALM OIL AND OLEIN BLENDS AT SLOW CRYSTALLIZATION (10 ¥ 10 MAGNIFICATION)
A, B and C are Blend B with 0.03, 0.06 and 0.09% lecithin, respectively; D, E and F are Blend E
with 0.03, 0.06 and 0.09% lecithin, respectively.
 EFFECTS OF EMULSIFIERS ON PO 11

olein. These high melting TAGs, especially the PPP, crystallize before the
other TAGs do, thus forming a new liquid and solid interface on the surface of
the crystal (De Man et al. 1989; Stauffer 1996; Zaki et al. 1997).
The increment in the SFCs of Blends J–L (0.1% SFC), E and F (6.7–6.9%
SFC), and B–D (7.9–9.1%) followed the same trend as the increment of PO in
Blends A–L at 0.03% lecithin. Except for Blends J–L, the increase in the SFCs
was significant with changes in PO content with 0.06 and 0.09% lecithin. As
shown in Table 1, insignificant SFCs in some of these blends indicates that
lecithin was also more effective in promoting crystal growth in one blend over
the other. Lecithin at 0.03% was more effective in promoting crystal growth in
Blends D than C, and F than E, and L than J and K. However, lecithin was more
of a crystal inhibitor at 0.06 and 0.09% in Blends A–H. Insignificant increase
in the SFCs of Blends J–L with 0.03–0.09% lecithin indicated that the low
content of SSS in the blends did not create sufficient crystals.
Viscosity, SFC Relationship with Lecithin
Viscosity is caused by the gelling effect of oil during the cooling process
(Faur 1996). The increase in viscosity during crystallization process may have
caused several implications, such as low heat conductivity and slowness in
heat transfer through the mass that led to the crystal growth selective only to
the high melting crystals (Faur 1996). The gelling properties due to the
molecular arrangements of the crystals had developed constraints on crystal
mobility and led to the minimum interactions between the crystals. Conse-
quently, it retarded the crystal growth leading to the low solid. However, slow
crystallization of PO and palm olein was contradictory with that of Faur
(1996). The increases of SFC during slow crystallization of Blends G–L with
0.06% lecithin (Fig. 4A) and D–L with 0.09% lecithin (Fig. 4B) were not
followed by an increase of viscosity. On the other hand, the reduction in the
viscosity was more of a general effect of lecithin (comparing Fig. 4A,B).
The viscosities of Blends D–L with lecithin at 0.03% were below 200 cps, and
the viscosities of Blends G–L with 0.06% lecithin were also below 200 cps
(Fig. 4A). However, the increase in trisaturated fatty acids content had caused
the increased in SFCs that was followed by the increase in viscosity of Blends
A–F (Fig. 4A) and A–C (Fig. 4B) for lecithin contents of 0.06 and 0.09%,
respectively.
The viscosity of the blends with lecithin was closely related to the crystal
dispersion in the liquid as shown in Fig. 3. Variation in the crowdedness of the
crystal distribution in Blend B with 0.03, 0.06 and 0.09% lecithin had viscosi-
ties of 394, 725 and 1,343 cps, respectively, as shown in Table 2. The viscosi-
ties of Blend E of the same lecithin content were 112, 384, and 186 cps, and
Blend G had viscosities of 101, 117 and 112 cps with the same lecithin content,
respectively.
12 M.S. MISKANDAR ET AL.

10 1200

A
8 1000

SFC
6 800
VISC

Viscosity (cps)
SFC (%)

4 600
R2 = 0.9869

2 400
R2 = 0.9757

0 200
A B C D E F G H I J K

-2 0
PO Blend with 0.06% Lecithin

10 2000

1800
B
8
1600
SFC
1400
6 VISC
Viscosity (cps)

1200
SFC (%)

4 1000

800
2
600

400
0
2
A B C D E F G H RI = 0.9543
J K 200
R2 = 0.962
-2 0
PO Blend with 0.09% lecithin

FIG. 4. SOLID FAT CONTENT (SFC) (%) AND VISCOSITY (VISC, CPS) AS A FUNCTION OF
PALM OIL (PO) AND PALM OLEIN BLENDS WITH 0.06% (A) AND 0.09% (B) LECITHIN AT
SLOW CRYSTALLIZATION
 EFFECTS OF EMULSIFIERS ON PO 13

Effects of STS on Crystal Behavior

STS in PO and olein blends was generally a crystal inhibitor, as shown
by its insignificant effect on the SFC except for few blends. Addition of STS
at 0.03% level into the fat Blends G–L did not significantly (P > 0.05) change
the SFC of the blends (except for Blends E, F and H) at the end of cycle 1
(Table 1). The SFC of Blends E, F and H had increased (P < 0.05) from 6.4
to 6.7, from 5.5 to 6.4 and from 1.6 to 2.6%, respectively. The increase in
SFC indicated that STS was a crystal promoter in these blends. As 0.06%
STS was added into the fat blends, no significant changes occurred on the
SFCs except for a few blends. The SFC decreased (P < 0.05) in A, B and H,
but an increase in F was noted. At 0.09% STS, the SFC decreased (P < 0.05)
in B, D, F and H. Table 1 shows that STS had promoted the crystal growth
of a few blends, but inhibited growth of most of the blends. These findings
are in agreement with those of Garti (1988) on the inhibition of the effect of
STS at 0.3% in sunflower oil margarine, although this study used lower
percentages of STS.

SFC, Viscosity Relationship with STS

SFC did not show a linear relationship with viscosity in blends with STS
(Fig. 5A,B). The results show that STS was only effective in lowering the
viscosities of the low-melting blends (Blends F–L). The viscosities were
maintained at below 200 cps in Blends G–L with 0.03% STS, F–L with 0.06%
STS (Fig. 5A) and G–L with 0.09% STS (Fig. 5B). It was also noted that
increase of viscosities was only at region of SFC > 4% in Blends A–F, A–C
(Fig. 5A) and A–F (Fig. 5B) with STS content of 0.03, 0.06 and 0.09%,
respectively.
Addition of emulsifier altered the chemical composition of PO and palm
olein blends, hence their physical properties. The interaction of the fatty acids
of the TAG and the acyl groups of the emulsifier was selective and led to
different rates of retardation and acceleration of crystal growth (Katsuragi
1999). Acceleration of crystal growth occurred when the emulsifiers were
adsorbed by the kinks of TAG molecules. STS seemed to have some of the acyl
ends’ preference toward other similar acyl ends belonging to the TAGs
forming the crystal, and were adsorbed to them, resulting in the retardation in
growth of that crystal type, but at the same time allowing other crystals to
continue growing and causing SFC to increase (Garti 1988; Katsuragi 1999).
The preference of the acyl ends of STS and TAG was manifested by the
homogeneous crystal distribution of even sizes in Blends B and E at 0.03%
STS, respectively (Fig. 6). It is suggested that STS preferred blends with
trisaturated fatty acids, and the interaction of acyl–acyl groups had enhanced
the formation of homogeneous crystal sizes. The low trisaturate content in
14 M.S. MISKANDAR ET AL.

10 2200

A 2000
8 1800
SFC
1600
6 VISC
1400

Viscosity (cps)
SFC (%)

1200
4
R2 = 0.9581 1000
800
2
600
R2 = 0.9423
0 400
A B C D E F G H I J K 200
-2 0
PO Blend with 0.06% STS

10 1600
B
1400
8
SFC 1200
Visc
6
1000
Viscosity (cps)
SFC (%)

4 800
R2 = 0.9562
600
2 R2 = 0.9838
400
0
200
A B C D E F G H I J K

-2 0
PO Blend with 0.09% STS

FIG. 5. SOLID FAT CONTENT (SFC) (%) AND VISCOSITY (VISC, CPS) AS A FUNCTION OF
PALM OIL (PO) AND PALM OLEIN BLENDS WITH 0.06% (A) AND 0.09% (B) SORBITAN
TRISTEARATE (STS) AT SLOW CRYSTALLIZATION

Blend G may have caused development of hardly visible crystals in Blend G

by normal microscope. The crystal dispersion micrograph showed that crystal
migration had occurred toward a particular nucleus with STS 0.06 and 0.09%
forming irregular crystal sizes. Blends with more PO content had more crystal
 EFFECTS OF EMULSIFIERS ON PO 15

A C

B D
FIG. 6. PHOTOMICROGRAPH OF EFFECT OF SORBITAN TRISTEARATE (STS) ON
CRYSTAL DISPERSION OF PALM OIL AND OLEIN BLEND AT SLOW CRYSTALLIZATION
(10 ¥ 10 MAGNIFICATION)
A and B are Blend B with 0.03 and 0.09% STS, respectively; C and D are Blend E with 0.03 and
0.09% STS, respectively.

formation and in the same manner more dense crystal formation occurred in
blends with 0.09% STS.
The viscosity of the blends with STS was closely related to the crystal
dispersion in the liquid as shown in Table 3 and Fig. 6, respectively. Variation
in the crowdedness of the crystal distribution in Blend B with 0, 0.03, 0.06 and
0.09% STS had viscosities of 773, 960, 1,506 and 1,166 cps, respectively, as
shown in Table 2. The crystals in Blend B (0% STS) formed into big agglom-
erates that separated from the liquid and led to the low viscosity. Blend B
(0.03% STS) produced small but evenly dispersed crystals with the lowest
viscosity. The blends with 0.09 and 0.06% STS had promoted big and closely
16 M.S. MISKANDAR ET AL.

TABLE 3.
THE EFFECTS OF SORBITAN TRISTEARATE (STS) ON THE VISCOSITY (CPS) OF PALM
OILAND OLEIN BLENDS BY CYCLE 1

Blend % STS

0 0.03 0.06 0.09

B 773 ⫾ 78Ac 960 ⫾ 160Ac 1506 ⫾ 170Cb 1166 ⫾ 15Bc

E 496 ⫾ 235Bb 554 ⫾ 64Bb 202 ⫾ 46Aa 416 ⫾ 27Bb
G 138 ⫾ 9Ba 122 ⫾ 9Aa 112 ⫾ 0Aa 122 ⫾ 9Aa

Mean of three determinations ⫾ SD.

A–C
Mean with different letters within each row is significantly different (P < 0.05).
a–c
Mean with different letters within each column is significantly different (P < 0.05).
Cycle 1, treatment temperatures and time of 80C (30 min) → 20C (80 min).

packed crystal agglomerates with a strong crystal network, leading to high

product viscosity. The viscosities of Blend E were 496, 554, 202 and 416 cps,
and Blend G had viscosities of 138, 122, 112 and 122 cps with the same STS
content, respectively. The trisaturate content of Blend E was lower than B;
thus, crystal network formation was lesser in strength and lower in viscosity
because of the dramatic increase in crystal size, but significantly decreased in
crystal number (Martini et al. 2002). The phenomenon suggested that as more
STS was present in the blend, nuclei formation took place much earlier during
cooling. The undesirable nuclei formation had enhanced crystallization,
leading to the formation of large crystal size. The concentration of these
crystals at limited numbers of centers was undesirable as crystal network
would not be sufficient to bind the liquid, thus causing oil to separate from the
solid. Table 3 shows that the viscosity in blends with 0.06 and 0.09% STS was
greatly reduced because of oil separation.

CONCLUSIONS

The Tm and Tc under slow crystallization conditions were sufficient for the
PO and olein to crystallize without an emulsifier. However, crystallization was
affected by the presence of either lecithin or STS. These emulsifiers, at specific
amount and temperature, lead to crystal promotion or retardation. Under static
condition, crystallization was toward the nearest nuclei. Heterogeneous crys-
tals were formed in blends with high trisaturates, and homogeneous crystals
were significantly distributed in blends with more liquid phase. The slow
crystallization process of PO and olein blends with 0.03% lecithin produced
homogeneous crystals with low SFC and viscosity for fluidized products.
 EFFECTS OF EMULSIFIERS ON PO 17

ACKNOWLEDGMENTS

M.S. Miskandar thanks the Malaysian Palm Oil Board for permission to
use the laboratory facilities. Special thanks to the Ministry of Science and
Technology Malaysia for financing the study and to Radzuan, Hanirah, Rosnah
and Nor Haya for providing technical assistance.

REFERENCES

AOCS 1989. Official and Tentative Methods of American Oil Chemist Society,
4th Ed., AOCS, Champaign, IL.
BLAUROCK, A.E. 1999. Fundamental understanding of the crystallization of
oils and fats. In Physical Properties of Fats, Oils, and Emulsifiers, (N.
Widlak, ed.) pp. 1–30, AOCS Press, Champaign, IL.
BUSFIELD, W.K. and PROSHOGO, N.P. 1990. Isothermal analysis of palm
stearin by DSC. J. Am. Oil Chem. Soc. 67, 171–180.
CHRYSAM, M.M. 1985. Table spreads and shortening. In Bailey’s Industrial
Oil and Fat Products, Vol 3 (T. Applewhite, ed.) pp. 42–83, John Wiley
and Sons, Inc., New York, NY.
DE MAN, L., DE MAN, J.M. and BLACKMAN, B. 1989. Polymorphic
behavior of some fully hydrogenated oils and their mixtures with liquid
oils. J. Am. Oil Chem. Soc. 66, 1777–1780.
FAUR, L. 1996. Fractionation technology. In Oils and Fats Manual, Vol 2 (A.
Karleskind, ed.) pp. 935–951, Lovoisier Publishing, Paris, France.
GARTI, N. 1988. Effects of surfactants on crystallization and polymorphic
transformation of fats and fatty acids. In Crystallization and Polymor-
phism of Fats and Fatty Acids (N. Garti and K. Sato, eds.) pp. 267–300,
Marcel and Dekker, Inc., New York, NY.
GARTI, N. 2002. Food emulsifiers: Structure-reactivity relationships, design,
and applications. In Physical Properties of Lipids (A.G. Marangoni and
S.S. Narine, eds.) pp. 265–340, Marcel and Dekker, Inc., New York, NY.
GHOTRA, B.S., DYAL, S.D. and NARINE, S.S. 2002. Lipid shortening: A
review. Food Res. Int. 35, 1015–1048.
JAHANIAVAL, F., KAKUDA, Y. and ABRAHAM, V. 2000. Oil-binding
capacity of plastic fats: Effects of intermediate melting point TAG. J. Am.
Oil Chem. Soc. 79, 389–394.
KAMEL, B.S. 1997. Emulsifiers. In Food Additive User’s Handbook
(J. Smith, ed.) pp. 170–201, Blackie Academic & Professional, London,
U.K.
KATSURAGI, T. 1999. Interactions between surfactants and fats. In Physical
Properties of Fats, Oils, and Emulsifiers (N. Widlak, ed.) pp. 211–219,
AOCS Press, Champaign, IL.
18 M.S. MISKANDAR ET AL.

KROG, N. and LAST, D. 1995. Palm oil based food emulsifiers. J. Mal. Oil
Sci. Technol. 4, 93–100.
MARTINI, S., HARRERA, M.L. and HARTEL, H.R. 2002. Effect of cooling
rate on crystallization behavior of milk fat fraction/sunflower oil blends.
J. Am. Oil Chem. Soc. 79, 1055–1062.
MISKANDAR, M.S., CHE MAN, Y.B., YUSOFF, M.S.A. and ABDUL
RAHMAN, R. 2002. Effect of emulsion temperature on physical prop-
erties of palm oil-based margarine. J. Am. Oil Chem. Soc. 79, 1163–1168.
MISKANDAR, M.S., CHE MAN, Y.B., YUSOFF, M.S.A. and ABDUL
RAHMAN, R. 2004. Palm oil crystallization: Effects of cooling time and
oil content. J. Food Lipids 11, 190–207.
NASH, N.H. and LAWRENCE, M.B. 1972. Food emulsifiers-science and art.
J. Am. Oil Chem. Soc. 49, 457–461.
O’BRIEN, R.D. 2000. Shortening technology. In Introduction to Fats and Oils
Technology, 2nd Ed. (R.D. O’Brian, W.E. Farr and P.J. Wan, eds.)
pp. 422–451, AOCS Press, Champaign, IL.
STAUFFER, C.E. 1996. Properties of emulsifiers. In Fats and Oils (C.E.
Stauffer, ed.) pp. 29–33, Eagen Press, St. Paul, MN.
SUGIMOTO, T., MORI, T., MANO, J., MUTOH, T., SHINOKI, Y. and
MATSUMARA, Y. 2001. Effect of fat crystallization on the behavior of
proteins and lipids at oil droplet surfaces. J. Am. Oil Chem. Soc. 78,
183–188.
ZAKI, M.S., NIK MERIAM, S. and SIVARUBY, K. 1997. Triacylglycerols
responsible for the onset of nucleation during clouding of palm olein. J.
Am. Oil Chem. Soc. 74, 1553–1558.