E

:
F
o
o
d
E
n
g
i
n
e
e
r
i
n
g
&
P
h
y
s
i
c
a
l
P
r
o
p
e
r
t
i
e
s
Ultrasonic Characterization of Pork Fat
Crystallization during Cold Storage
Edith Corona, Jose V. Garca-Perez, Juan V. Santacatalina, Sonia Ventanas, and Jose Benedito
Abstract: In this work, the feasibility of using ultrasonic velocity measurements for characterizing and differentiating
the crystallization pattern in 2 pork backfats (Montanera and Cebo Iberian fats) during cold storage (0 C, 2 C, 5 C,
7 C, and 10 C) was evaluated. The fatty acid prole, thermal behavior, and textural properties (hardness) of fat were
also determined. Both fats became harder during cold storage (average hardness increase for both fats, 11.5 N, 8 N, and
1.8 N at 0, C 2 C, and 5 C, respectively), showing a 2-step pattern related with the separate crystallization of the
different existing triacylglycerols, which was well described using a modied Avrami equation (explained variance >
99%). Due to a greater content of saturated triacylglycerols, Cebo fat (45.1%) was harder than Montanera (41.8%). The
ultrasonic velocity followed a similar 2-step pattern to hardness during cold storage, being found an average increase for
both fats of 184, 161, and 150 m/s at 0 C 2 C, and 5 C, respectively. Thus, ultrasonic measurements were useful
both to characterize the textural changes taking place during cold storage and to differentiate between fats with different
composition.
Keywords: crystallization, fat, texture, mathematical modeling, ultrasound
Practical Application: The cold storage of dry-cured meat products during their distribution and retail sale exert an
important effect on their textural properties and consumers acceptance due to the crystallization of the fat fraction,
which is greatly inuenced by the type of fat. In this work, a nondestructive ultrasonic technique was used to identify
the textural changes provoked by the crystallization during cold storage, and to differentiate between fats, which could
be used for quality control purposes.
Introduction
To extend the shelf life of dry-cured meat products, preservation
technologies, such as freezing and refrigeration, are used during
distribution and retail sale (Cilla and others 2006). Nevertheless,
the storage time and fat content exerted a notable effect on the
textural properties of dry-cured meat products due to the crystal-
lization of the fat and the increase in the solid fat content (SFC)
at low temperatures (Campos and others 2002; Garca-Esteban
and others 2004; Rubio and others 2007; Santacatalina and others
2010). Therefore, monitoring the changes that take place during
the cold storage of fat is relevant because the texture of dry-cured
products is directly related to these variations. In addition, previous
studies (Ventanas and others 2008a; Fuentes and others 2010) have
reported the effect of fat content, composition, and state on the
release of volatile compounds which affect the perception of the
sensory attributes (odor and avor). This issue is largely noticeable
in dry-cured products of high-sensory quality and wide consumer
acceptance,such as those obtained from Iberian pig (Ventanas and
others 2005).
In Portugal and Spain, the Iberian pig is an autochthonous breed
traditionally reared and fed with a diet based on acorns and grass
(Montanera) and/or with concentrate feeds (Cebo; Tim on and
MS 20131485 Submitted 10/18/2013, Accepted 1/26/2014. Authors Corona,
Garca-P erez, Santacatalina, and Benedito are with Grupo de An alisis y Simulaci on
de Procesos Agroalimentarios, Dept. Tecnologa de Alimentos, Univ. Polit` ecnica de
Val` encia, Cam de Vera s/n, E46022, Valencia, Spain. Author Ventanas is with
Tecnologa de los Alimentos, Dept. de Producci on Animal y Ciencia de los Alimentos,
Facultad de Veterinaria, Univ. de Extremadura, Avd.de la Universidad s/n, 10071,
C aceres, Spain. Direct inquiries to author Garcia-Perez (E-mail: jogarpe4@tal.upv.es).
others 2001a). The inuence of the rearing system leads to modi-
cations in the fatty acid prole of the raw material, thus affecting
the sensory characteristics of the derived meat products (Carrapiso
and others 2003; Ni noles and others 2007). In this regard, Montan-
era system allows to obtain tissues with higher levels of unsaturated
fatty acids, particularly of oleic acid, than those derived from pigs
fed on Cebo system (Tim on and others 2001b; Petr on and others
2004), which affects sensory characteristics (brightness, oiliness,
aroma, and juiciness; Ruiz-Carrascal and others 2000), as well
as crystallization pattern. Therefore, dry-cured products derived
from Montanera Iberian pigs show a much higher overall sensory
quality and market prices are much higher than those produced
from Cebo Iberian pigs. However, although the Iberian meat
product market is regulated (BOE 2007), sometimes, the products
derived from Cebo pigs are labeled as Montanera. Therefore, it
is important to nd technologies that can adequately identify and
classify the different products in function of the rearing system
used to prevent possible fraud and maintain quality standards.
Different techniques have been used for fat characterization and
to monitor changes during fat crystallization (Cerdeira and others
2004), such as instrumental textural analysis, differential scanning
calorimetry (DSC), X-ray diffraction, pulsed nuclear magnetic
resonance, and polarized light microscopy. However, the search
for nondestructive techniques, like ultrasound, that could be used
to monitor the crystallization pattern of complex fats during cold
storage is of great interest to the industry. Among other advan-
tages of ultrasound, it is rapid, low cost, and easy to be auto-
mated online (Mulet and others 1999; Yucel and Coupland 2011).
Low-intensity ultrasound has been used to assess the textural prop-
erties of many foods, such as Mahon Cheese (Benedito and others
C
2014 Institute of Food Technologists
R
E828 Journal of Food Science
r
Vol. 79, Nr. 5, 2014 doi: 10.1111/1750-3841.12410
Further reproduction without permission is prohibited
E
:
F
o
o
d
E
n
g
i
n
e
e
r
i
n
g
&
P
h
y
s
i
c
a
l
P
r
o
p
e
r
t
i
e
s
Ultrasonics monitoring crystallization . . .
2000) and meat-based products (Sobrassada; Llull and others 2002).
Ultrasonic measurements were also useful to characterize and clas-
sify subcutaneous fat from dry-cured Iberian hams with different
genetics and rose on different feeding systems (Ni noles and others
2008). Moreover, the increase in the SFC was monitored during
crystallization of rened lard (Santacatalina and others 2010) and
milk and vegetable fats (Singh and others 2004; Martini and others
2005a, 2005b) using ultrasonic measurements. Maleky and others
(2007) also used low-intensity ultrasound to quantify mechanical
and structural characteristics of lipids. However, to our knowl-
edge, there is not any previous work in the literature regarding
the use of ultrasonics to characterize the textural changes during
crystallization in nonrened fats. Thereby, the aim of this work
was to test the feasibility of low-intensity ultrasound for both the
characterization of the textural changes in fats with different com-
position during cold storage and the discrimination between fat
types according to the crystallization pattern.
Materials and Methods
Raw material
Experiments were carried out using 2 Iberian subcutaneous fats:
Montanera and Cebo. In both fats, a homogeneous sample of 5000
g was obtained from the backfat of 10 Iberian pigs, approximately
0.5 kg of fat per animal. For each fat type, the fat was separated
from the skin before the experiments, milled, and homogenized
by melting at 70 C for 30 min and nally stored at 2 C.
Fatty acid composition
Fatty acid methyl esters (FAMEs) of both fats were prepared
by acidic-transesterication in the presence of sulphuric acid (5%
sulphuric acid in methanol; Sandler and Karo 1992). Lipids were
previously extracted by using the Folch and others (1957) proce-
dure. FAMEs were analyzed by a gas chromatograph (HP-5890A,
Hewlett Packard, Palo Alto, Calif., U.S.A.) equipped with a ame
ionisation detector. Separation was carried out on a polyethy-
lene glycolTPA modied fused silica semicapillary column (30
m long, 0.53 mm id, 1 m lm thickness) maintained at 225 C.
Injector and detector temperatures were 230 C. Carrier gas was
nitrogen at a ow rate of 1.8 mL/min. Individual FAMEs peaks
were identied by comparing their retention times with those of
standards (Sigma-Aldrich Corp., St. Louis, Mo., U.S.A.). Results
are expressed as percentage of the total fatty acids analyzed. Eight
repetitions per fat were performed.
Differential scanning calorimetry
The thermal behavior of the samples was monitored using DSC
(DSC5200CU; Seiko Instruments, Inc., Torrance, Calif., U.S.A.).
Each fat type was analyzed at least in triplicate. Approximately 18
to 20 mg of dry fat were introduced into a standard aluminum
DSC crucible and then hermetically sealed and weighed on a bal-
ance ( 0.01 mg, ER-182A, AND, Tokyo, Japan). The sample
was previously dried in an oven at 105 C until constant weight
(approximately 24 h), to avoid the inuence of water. An empty
and hermetically sealed aluminum DSC crucible was used as the
reference material and liquid nitrogen was used as the cooling
uid. The sample was heated from 25 C to 70 C at 5 C/min.
Next, the sample was tempered at 70 C for 1 min to destroy any
crystal memory. Afterwards, the sample was cooled from 70 C
to 50 C using a wide range of cooling rates (0.2, 0.5, 1, 5,
and 10 C/min). The temperature range and cooling rates were
not chosen to reproduce the experiments conducted to study the
textural and ultrasonic velocity variations during cold storage (sec-
tions 2.4 and 2.5), but to evaluate if the fractions of triglycerides
with different crystallization properties could be identied and
also to assess the discriminating capacity of DSC to differentiate
between Cebo and Montanera fat. Following this procedure, DSC
curves were obtained for the cooling of both fats at the 5 different
cooling rates tested. The peak maximum temperature (T
max
, C),
onset temperature (T
o
, C) and latent heat (H, J/g) were com-
puted from the thermograms (Cebula and Smith 1992).
Instrumental texture analysis
The textural changes of the fat samples during cold storage
(0 C to 5 C) were analyzed using a universal texture analyzer
(TA-XT2i; Stable Micro Systems, Surrey, England) linked to a
computer for data acquisition and processing. To perform textural
measurements, the fat was previously melted at 70 C for 30 min
in a vacuum oven (EV50, Raypa Vacuterm, Terrasa, Spain) and
then placed in Petri dishes (volume of fat 60 mL, layer thickness
10 mm) and manually stirred until reaching 50 C. Petri dishes
and the textural analyzer were placed in the same temperature-
controlled chamber ( 0.1 C; AEC330R, Infrico, Barcelona,
Spain), which allowed the temperature of the samples to be kept
constant during the instrumental test. Three storage temperatures
were tested: 0 C, 2 C, and 5 C. These temperatures were
chosen as they are common for storage, retail sale, and distribution
of dry-cured meat products. From the rst hour of storage, one
Petri dish was measured every hour for approximately 5 d, using
a conical probe (Plexiglas, angle 40, penetration distance 5 mm
and crosshead speed 1 mm/s). For each time, 5 penetrations were
carried out on the sample placed in each Petri dish. Finally, the
maximum force (MF, N) was computed from the force versus
distance records and averaged for the 5 measurements. Therefore,
the MF was considered in this work to evaluate the changes in
hardness during fat crystallization.
Ultrasonic set-up and measurements
The ultrasonic experimental set-up (Figure 1) consisted of an
aluminum cylindrical ultrasonic cell (diameter 17.8 mm, length 25
mm) where the sample was introduced. A narrow band piezoelec-
tric transducer (1 MHz, 0.5 crystal diameter, A303S, Panametrics,
Walthman, Mass., U.S.A.) was inserted in each side of the cell, the
path length being 25 mm. A thermocouple K-type was used to
measure the sample temperature; it was placed in the center of the
cell, which was kept in a temperature-controlled bath (0.1 C;
Frigiterm-30, P-selecta, Barcelona, Spain). The ultrasonic pulser-
receiver instrument (5058PR, Panametrics) supplied the electrical
spike pulse to the emitter-transducer generating the ultrasonic
wave, which propagated through the sample and was received by
the receivertransducer. Then, the signal was ltered (0.3 MHz
LP5) and amplied (40 dB) by the pulser-receiver and thereafter,
captured and digitized (100 MS/s) by a data acquisition card (PCI
5112, National Instruments, Austin, Tex., U.S.A.) installed in a
PC. Specic software was programmed in Visual Basic for the
determination of the ultrasonic velocity, taking the time of ight
and the distance between transducers into account. Time of ight
is dened as the time the ultrasonic wave needs to go through the
sample, whereas the distance between the transducers was calcu-
lated using distilled water. The accuracy of the ultrasonic velocity
measurement was of 3 m/s.
Vol. 79, Nr. 5, 2014
r
Journal of Food Science E829
E
:
F
o
o
d
E
n
g
i
n
e
e
r
i
n
g
&
P
h
y
s
i
c
a
l
P
r
o
p
e
r
t
i
e
s
Ultrasonics monitoring crystallization . . .
The experiments were carried out on samples of fat that
had previously been stored at 2 C, then heated at 70 C for
30 min in a vacuum oven (EV50; Raypa Vacuterm, Terrassa,
Spain) to ensure the complete destruction of the fat crystals.
Thereafter, the sample was manually stirred until a temperature of
50 0.1 C was reached and placed into the ultrasonic cell
(Figure 1). Once the sample was introduced into the ultrasonic
cell, it was placed in a bath at 45 Cfor 10 min to assure tempering.
Then the ultrasonic cell was introduced into the temperature-
controlled bath set to the experimental temperature (0 C, 2 C,
5 C, 7 C, and 10 C). Once the sample reached the bath temper-
ature, the isothermal storage experiment began and the ultrasonic
velocity was measured every 10 min during storage times ranging
between 100 and 480 h (depending on the storage temperature).
Three replicates were carried out for each temperature and type
of fat.
As the ultrasonic set-up allowed long periods of cold stor-
age to be monitored in an automated and nondestructive way,
a wide range of storage temperatures were considered. Therefore,
tests were performed for conventional storage temperatures (0 C,
2 C, and 5 C) and temperatures on the limits of cold storage
(7 C and 10 C).
Modeling of ultrasonic velocity and statistical analysis
Several mathematical equations have been proposed to describe
the isothermal crystallization pattern of fats (Foubert and others
2004; Santacatalina and others 2010). In this work, the modied
Avrami equation (Eq. 1), was used to analyze how ultrasonic ve-
locity changes during the isothermal portion of the crystallization
process:
V(t ) = V
0
+a (1 exp (k t
n
)) (1)
where V is the ultrasonic velocity (m/s) at time t (s), V
0
the ultra-
sonic velocity at the beginning of the isothermal crystallization, a
the ultrasonic velocity at innite time (m/s), k is the crystallization
rate (s
n
) dependent on the crystallization temperature and n the
Avrami exponent (dimensionless).
For 2-stage crystallization, which is common in complex fats,
Santacatalina and others (2010) proposed the following expression
based on Eq. 1:
V = A
1
Avr
1
+ A
2
Avr
2
(2)
where Avr
1
and Avr
2
are the modied Avrami equations (Eq. 1)
for the rst and second crystallization stages, respectively, and A
1
and A
2
can be calculated using Eq. 3 and 4:
A
1
=

exp

if t >
1 if t

(3)
A
2
= 1 A
1
(4)
where is the time when the second crystallization stage starts
and a parameter that controls the curvature index at the junction
of the 2 fuzzy sets (represented by the 2 Avrami equations).
The accuracy with which the model ts the experimental data
was evaluated by the percentage of explained variance (percentage
VAR) and the root mean square error (RMSE, m/s). The analysis
of variance (ANOVA) was carried out to determine the signif-
icant inuence of the rearing system (Montanera and Cebo) on
the fatty acid prole, thermal properties, ultrasonic velocity, and
texture. The ANOVAs were performed using the Statgraphics
R
Centurion XVI statistics software and Fishers least signicant dif-
ference intervals were computed at a 95% condence level.
Results and Discussion
Fatty acid prole
The fatty acid prole in Montanera and Cebo fats are shown in
Table 1, where the content of saturated (SFA), monounsaturated
(MUFA), and polyunsaturated fatty acids (PUFA) can be com-
pared for both fats. The results show that there was a signicant
difference (P < 0.05) between the SFA content of both fats, being
higher for Cebo (45.1%) than for Montanera (41.8%). SFA con-
tent is affected by the rearing system, which involves lower SFA
percentages in Montanera fat from pigs fed on a more extensive,
natural diet based on acorns and grass than Cebo animals (Ni noles
and others 2007). The main SFA in both fats were palmitic (C16)
and stearic (C18) acids, and the differences in SFA content were
mainly due to the differences in stearic acid. These results agree
with those previously found by Petr on and others (2004), who re-
ported higher proportions of palmitic and stearic acids in fat from
Iberian pigs fed on concentrate feed.
On the other hand, in the case of the unsaturated fatty acid con-
tent (MUFA + PUFA), Montanera showed a signicantly (P <
0.05) higher content than Cebo (58.2% versus 54.9%; Table 1).
For both MUFAand PUFAfractions, as expected, Montanera pre-
sented a signicantly (P < 0.05) higher content than Cebo, which
Figure 1Ultrasonic system for monitoring
isothermal crystallization.
E830 Journal of Food Science
r
Vol. 79, Nr. 5, 2014
E
:
F
o
o
d
E
n
g
i
n
e
e
r
i
n
g
&
P
h
y
s
i
c
a
l
P
r
o
p
e
r
t
i
e
s
Ultrasonics monitoring crystallization . . .
Table 1Fatty acid composition (%) of Montanera and Cebo
backfats.
Cebo Montanera
C12 0.06 0.04
a
0.070 0.014
b
C14 1.51 0.09
a
1.67 0.08
a
C16 30.1 0.8
a
29.1 0.9
a
C16:1 2.03 0.04
a
1.98 0.04
a
C17 0.35 0.02
a
0.31 0.02
a
C17:1 0.301 0.009
a
0.260 0.006
b
C18 12.7 0.5
a
10.46 0.3
b
C18:1 (n-9) 42.3 0.5
a
45.07 1.08
b
C18:1 (n-7) 3.37 0.11
a
2.58 0.13
b
C18:2 4.2 0.3
a
6.3 0.3
b
C18:3 (n-3) 0.24 0.12
a
0.41 0.09
a
C20 0.4 0.2
a
0.246 0.112
a
C20:1 1.20 0.09
a
0.7 0.03
b
C20:2 0.44 0.13
a
0.32 0.04
a
C20:3 (n-6) 0.17 0.09
a
0.12 0.02
a
C20:4 0.40 0.12
a
0.20 0.09
a
C20:3 (n-3) 0.31 0.11
a
0.15 0.05
a
C24:0 0
a
0.021 0.002
a
SFA 45.13 0.14
a
41.83 0.16
b
MUFA 49.15 0.09
a
50.62 0.25
b
PUFA 5.72 0.14
a
7.55 0.08
b
MUFA + PUFA 54.87 0.07
a
58.17 0.19
b
SFA, Saturated fatty acid; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated
fatty acid.
Average values standard deviation are shown from 8 replicates. Different letters (a and
b) in the same row denote a signicant difference (P < 0.05).
is mainly linked to the difference in the oleic (C18:1; 45.07%
forMontanera and 42.3% for Cebo) and linoleic acid contents
(C18:2; 6.3% for Montanera and 4.2% for Cebo). The high con-
tent of oleic acid in both fats is explained by the feeding system.
As previously mentioned, Montanera animals are mainly fed on
acorns and grass, which are rich sources on this fatty acid (Ni noles
and others 2007). In the case of Cebo animals, the concentrate
feeds used also contain a high content of oleic acid, aiming to
simulate the supply provided by acorns.
The fatty acid composition of Montanera and Cebo fats found
in this work was very similar to others previously reported
(Carrapiso and others 2003; Ventanas and others 2008a). The
fatty acid prole has a great inuence on the sensory and textu-
ral properties of meat products (Ruiz-Carrascal and others 2000).
Although triacylglycerols (TGs) have not been identied in this
work, it could be assumed that the level of unsaturation in TGs
is directly linked to the fatty acid composition. Consequently, the
crystallization pattern of Iberian Montanera and Cebo fats is also,
to a certain extent, related with the fatty acid composition. This
issue is addressed in the following section.
Thermal behavior characterization
The thermal behavior of Montanera and Cebo subcutaneous fat
was analyzed by recording DSC curves at different cooling rates
of 0.2, 0.5, 1, 5 and 10 C/min. Different cooling rates were used
with the aim of identifying the crystallization pattern of both fats
and nding the rate that provides the best differentiation between
them.
Similar DSC curves were found for both fats, showing 2 main
peaks (A and B) for every cooling rate (Figure 2). This fact has
already been observed and linked to the crystallization of different
TGs in fats with a complex chemical composition (Le Mestre and
others 1984). Peak A may be related with the crystallization of
the most highly saturated TGs, whereas peak B may be linked
to the crystallization of the most unsaturated TGs. In the case of
Montanera (Table 2), the onset temperature (T
o
) of peak A ranged
from 20.5 (at 0.2 C/min) to 5.2 C (at 10 C/min) and for peak
B, from 1.6 (at 0.2 C/min) to -26.0 C (at 10 C/min). The
T
max
for Montanera fell from 18.1 C to 5.0 C for peak A and in
peak B, from 5.8 C to 34.5 C when the rate increased from
0.2 to 10 C/min. The effect of the cooling rate has previously
been studied in lard (Campos and others 2002; Kalnin and others
2005)using the combined DSC and X-ray diffraction techniques.
These studies also found that the higher the cooling rate, the lower
the T
o
and T
max
, which is coherent with the results obtained in
this work.
Both peaks (A and B) were broadened and displaced to lower
temperatures when the cooling rate increased (Figure 2; Table 2).
These results are explained by a kinetic limitation due to the short
crystallization time at high cooling rates (Cebula and Smith 1991).
In the case of peak displacement, for the fastest rates, only a very
short time is needed to crystallize at a given temperature, forc-
ing the TGs to crystallize at lower temperatures, a phenomenon
known as supercooling. The short crystallization time also involves
peak broadening, which appears because TGs with similar crys-
tallization temperatures are not able to crystallize simultaneously
due to a time limitation.
For both A and B peaks, it could be stated that Montanera pre-
sented lower T
o
and T
max
than Cebo. This fact may be observed
in the thermograms by the displacement of the peaks for both fats
and could be linked to the higher level of unsaturation of TGs in
Montanera. In the case of peak A, Cebo showed a signicantly
(P < 0.05) higher T
o
than Montanera: the faster the cooling rate
(T
o
= 3.4 C at 0.2 C/min and 9.7 C at 10 C/min), the
greater the difference (T
o
). For the same peak, the T
max
was
signicantly higher (P < 0.05) for Cebo than Montanera at the
lowest cooling rates (from 0.2 to 1 C/min), but at 5 C/min, the
difference disappeared and at 10 C, Montanera presented a signif-
icantly (P < 0.05) higher T
max
than Cebo. In the case of peak B,
T
o
, and T
max
values were signicantly (P < 0.05) higher for Cebo
than Montanera (average T
o
= 6.6 C and T
max
= 5.7 C),
with the exception of the T
max
at 0.2 C/min. Thermograms of
Montanera and Cebo fats also differed in the area (latent heat of
crystallization) of peaks A and B (Table 2). For peak A, the latent
heat was signicantly (P < 0.05) higher in Cebo than Montanera,
which could be expected given the higher level of saturation in
Cebo TGs. Peak B was observed to behave in the opposite way,
thus, the higher level of unsaturation in Montanera than in Cebo
involved a signicantly (P < 0.05) higher latent heat for all cool-
ing rates. Regarding the discrimination capacity of the different
parameters, the H of peak B is the one that provided the biggest
difference between Cebo and Montanera fat (Table 2). Moreover,
the discrimination ability was very similar for the different cooling
rates, being at 1 C/min where the maximum differences in H
of peak B for Cebo and Montanera fats were found. Himawan and
others (2006) reported that the crystallization pattern is very sen-
sitive, even to small differences in fatty acid composition, which
explained the signicant (P < 0.05) differences observed between
Montanera and Cebo fats. It should be pointed out that, in this
study, signicant (P < 0.05) differences in the fatty acid compo-
sition were found between Montanera and Cebo fats. However,
through the use of specic fat sources in the animals concentrate
feed, it might be possible to steer the Cebo fats in such a way as
to mimic the composition of Montanera fat (Ventanas and oth-
ers 2008b), which would hinder the differentiation between both
rearing systems using DSC.
Vol. 79, Nr. 5, 2014
r
Journal of Food Science E831
E
:
F
o
o
d
E
n
g
i
n
e
e
r
i
n
g
&
P
h
y
s
i
c
a
l
P
r
o
p
e
r
t
i
e
s
Ultrasonics monitoring crystallization . . .
Characterization of the textural measurements during cold
storage
The textural properties of Montanera and Cebo fats were mon-
itored during cold storage at different temperatures (0 C, 2 C,
and 5 C). The maximum penetration force (MF) (Figure 3) was
measured due to its direct relationship with fat hardness. Before
the isothermal experiments were conducted, the fat samples were
tempered from 50 C to the storage temperature. During that
tempering process, part of the fat corresponding to the most satu-
rated TGs had been crystallized and therefore the initial MF (t =0)
corresponded to the texture of the partly crystallized sample. Thus
once tempering nalized and for both fats, the penetration force
increased during the isothermal storage for each of the experimen-
tal temperatures tested. For experiments carried out at 0 C and
2 C, the curves increased steeply twice during storage. The rst
increase took place during the rst hours of storage; this fact could
be linked to the crystallization of the most highly saturated TGs
found by DSC (peak A) which still remained liquid at the begin-
ning of the isothermal portion of the crystallization experiments.
Because the onset temperature of peak A was in all cases above
2 C (Table 2), part of the TGs of this peak should be crystallized
before the isothermal crystallization begun. Once the most highly
-0.5
0.5
1.5
2.5
3.5
4.5
5.5
-50 -30 -10 10 30
H
e
a
t

F
l
o
w

(
m
W
)
T (C)
Montanera
Cebo
A
A
B
B
1 C/min
-0.2
-0.1
0.0
0.1
0.2
0.3
-50 -30 -10 10 30
H
e
a
t

F
l
o
w

(
m
W
)
T (C)
Montanera
Cebo
0.2 C/min
A
B
B
A
0.0
2.0
4.0
6.0
8.0
10.0
-50 -30 -10 10 30
H
e
a
t

F
l
o
w

(
m
W
)
T (C)
Montanera
Cebo
5 C/min
A
A
B
B
-0.2
0.2
0.6
1.0
1.4
1.8
-50 -30 -10 10 30
H
e
a
t

F
l
o
w

(
m
W
)
T (C)
Montanera
Cebo
A
A
B
B
0.5 C/min
0.0
1.5
3.0
4.5
6.0
7.5
9.0
10.5
12.0
-50 -30 -10 10 30
H
e
a
t

F
l
o
w

(
m
W
)
T (C)
Montanera
Cebo
A
A
B
B
10 C/min
Figure 2DSC thermograms of Montanera and Cebo fats at different cooling rates: 0.2, 0.5, 1, 5, and 10 C/min.
E832 Journal of Food Science
r
Vol. 79, Nr. 5, 2014
E
:
F
o
o
d
E
n
g
i
n
e
e
r
i
n
g
&
P
h
y
s
i
c
a
l
P
r
o
p
e
r
t
i
e
s
Ultrasonics monitoring crystallization . . .
Table 2Thermal properties of Montanera and Cebo backfats obtained from DSC using different cooling rates: peak maximum
temperature (T
max
), onset temperature (T
o
) and latent heat (H).
T
max
(C) T
o
(C) H (J/g)
Cooling rate (C/min) Cebo Montanera Cebo Montanera Cebo Montanera
Peak A
0.2 21.2 1.1
a
18.1 1.3
b
23.9 0.3
a
20.5 0.6
b
24.1 1.1
a
18.3 1.3
c
0.5 19.7 1.6
a
14 2
c
21.4 1.1
b
18.2 1.2
c
25.2 0.7
a,b
17.5 1.0
c
1 18.5 0.6
b
9.4 0.5
d
20.1 0.8
b
15.2 1.8
d
24.1 0.6
a
16.2 0.9
d
5 6.4 1.1
d,e
7.6 0.5
d
16.7 0.6
c
10.5 1.0
e
25.2 1.1
a,b
18.3 0.6
c
10 1.7 0.5
f
5.0 0.7
e
14.7 1.1
d
5.2 0.7
f
23.3 1.3
a
13.5 0.5
e
Peak B
0.2 5.9 0.6
a
5.8 1.5
a
9.2 0.6
a
1.6 0.6
b
30.5 0.7
a
42.2 0.7
b
0.5 5.0 0.7
a
10.6 1.1
b
4.1 1.3
c
7.0 0.8
d
31.5 1.3
a
42.0 0.8
b
1 10.1 0.4
b
14 1.6
c
6.0 1.3
d
8.8 0.4
d,e
28.2 0.3
c
41.8 1.1
b
5 19 1.2
d
29.1 0.8
e
10.4 1.3
e
20.3 0.6
f
29.8 0.9
c,a
37.8 1.1
d
10 31.1 0.8
e
34.5 1.4
f
23.3 1.8
g
26.01 0.8
h
12.2 1.2
f
25.2 0.8
e
Average values standard deviation are shown from 3 replicates.
Superscript letters ah show homogeneous groups established from least signicant difference (LSD) intervals (P < 0.05) for each parameter (T
max
, T
o
, and H) and peak (A and B).
saturated TGs were fully crystallized, the MF remained constant
for a few hours, implying that cold storage progressed without
changes in the solid/liquid ratio. The duration of the constant
force period was dependent on storage temperature, showing that
the higher the storage temperature, the longer the constant period.
Thus, for Cebo fat, the constant period was extended to 35 h at
2 C, but only to 15 h at 0 C. Campos and others (2002) observed
that, when the fat system is rapidly chilled, the nucleation rate is
high and the crystallization takes place in a short period of time.
Finally, the second steep increase in the penetration force occurred
due to the crystallization of the most unsaturated TGs, which was
related to peak B found in DSC, as previously explained in section
3.2. The second increase also ended in a constant penetration force
period. It should be commented on that, in the case of samples
0
2
4
6
8
10
12
14
16
18
20
0 20 40 60 80 100 120
M
F

(
N
)
Time (h)
0 C
2 C
5 C
Model
A
0
2
4
6
8
10
12
14
16
18
20
0 20 40 60 80 100 120
M
F

(
N
)
Time (h)
0 C
2 C
5 C
Model
B
Figure 3Changes and modeling of the
maximum penetration force (MF) at
different temperatures of Montanera (A)
and Cebo (B) fats during cold storage.
Average values Standard deviation are
plotted from 5 replicates.
Vol. 79, Nr. 5, 2014
r
Journal of Food Science E833
E
:
F
o
o
d
E
n
g
i
n
e
e
r
i
n
g
&
P
h
y
s
i
c
a
l
P
r
o
p
e
r
t
i
e
s
Ultrasonics monitoring crystallization . . .
stored at 5 C, the most unsaturated TGs did not crystallize during
the 120 h of storage.
The hardness of the fat also varied in function of the cold storage
temperature. Thus, the lower the storage temperature, the higher
the penetration force. At 0 C, Cebo fat showed a MF close to 17 N
whereas at 2 C, it was of only 10 N (Figure 3B). At 5 C, the MF
was even lower (4 N) than at 2 C, due to the fact that only the most
highly saturated TGs crystallized and therefore, the solid/liquid
ratio was lower than at 2 C and 0 C. Montanera fat was observed
to behave similarly. The rapid crystallization could be the reason
of the hardness increase at low storage temperatures. In this regard,
Campos and others (2002) observed that lard crystallized rapidly
was harder than lard crystallized slowly and had a higher SFC.
Moreover, its solid state was in a more metastable polymorphic
form (). Upon slow crystallization, lard had a lower SFC and
its solid state was in a more stable polymorphic form (). These
authors found that the microstructure was also different between
the fast and slow crystallized fat. When lard crystallized rapidly,
crystallites were numerous and small, whereas a smaller number of
larger crystallites were observed when crystallized slowly (Campos
and others 2002).
When comparing the different penetration forces of Montan-
era and Cebo fats during storage at the same temperature (Fig-
ure 3), the results showed that Cebo was harder than Montanera
fat. Thus, the constant penetration force values after the rst in-
crease were higher for Cebo (8 N) than Montanera fat (2.5 N) at
0 C. This fact may be related with the higher content
of saturated TGs in Cebo fat (Petr on and others 2004),
which increases the solid/liquid ratio, resulting in a harder
texture (larger MF). When the most unsaturated TGs are
fully crystallized, which coincides with the end of the sec-
ond increase in the penetration force, Cebo fat still reached
a higher MF than Montanera (16 versus 8 N at 0 C).
This fact may be explained by considering the characteristic crys-
tallization pattern of saturated and unsaturated TGs. Saturated fatty
acids can easily align themselves to form a compact mass because
they are relatively linear molecules (Himawan and others 2006).
The unsaturated fatty acids have kinks in their aliphatic chains,
causing a zigzag bending, which has implications for crystal pack-
aging. This fact makes more unsaturated TGs have a less dense
structure, with lower melting temperatures than saturated TGs
with the same chain length (Himawan and others 2006). There-
fore, the higher level of saturation in Cebo fat implies a more
compact fat crystal packaging, resulting in a harder texture (larger
MF) during crystallization than Montanera fat. In addition to the
fat composition, the differences in the textural parameters could
also be attributed to differences in the structure of the samples.
In this regard, after only a short period of time, the use of the
Montanera rearing system leads to a signicant increase in the fat
content of the animal. As a result, the connective tissue is expected
to be less structured and consistent than that of intensively reared
animals (Ni noles and others 2008).
Monitoring of ultrasonic velocity during cold storage
Ultrasonic monitoring was performed during the isothermal
crystallization at different storage temperatures (0 C, 2 C,
5 C, 7 C, and 10 C) for both Montanera and Cebo fats. The
ultrasonic velocity measurements started when the temperature
of the sample reached the storage temperature, thus, the portion
of the crystallization under study always followed an isothermal
pattern. Before the isothermal experiments were conducted, the
fat samples were tempered from 45 C to the storage temperature.
During that tempering process, part of the fat corresponding to
the most saturated TGs had been crystallized and therefore the ini-
tial velocity (V
0
) corresponded to that of the partially crystallized
sample.
Figure 4 shows the 2-step sigmoid curves of ultrasonic veloc-
ity versus crystallization time. A similar behavior was observed
for the change of textural parameters during storage (Figure 3).
Thus, ultrasonic velocity increased with time, showing a pattern
with 2 steep increases (Figure 4). As already reported in section
3.3, the crystallization of the TGs shown in Figure 2 brings about
the increase in the fat solid/liquid ratio, which results in a harder
sample with a higher elastic modulus where ultrasound waves
travel faster than in samples with a higher liquid content. The
ultrasonic velocity in a solid medium is related to the Young or
elastic modulus (Martini 2013) and therefore to the textural prop-
erties of the material. The ultrasonic velocity has been related
with the textural properties of many foods, such as cooked car-
rots (Nielsen and Martens 1997), avocado (Mizrach and Flitsanov
1999), meat products (Llull and others 2002; Corona and others
2013), or cheese (Benedito and others 2000). In these works, the
increase in the ultrasonic velocity was related to the increase of
the samples hardness, for example during cheese maturation, an
accurate relationship between hardness and ultrasonic velocity was
found. Therefore, the increase of the ultrasonic velocity during
fat crystallization found in this work is expected to be linked to
a harder fat as shown in Figure 3(A) and (B). Singh and others
(2004) also found an increase of the ultrasonic velocity when the
SFC of fats increased as a consequence of the crystallization. It
is important to highlight that the increase of the ultrasonic ve-
locity and the MF are not coincident in time (Figure 3 and 4).
The measurement of both MF and ultrasonic velocity began when
the storage temperature was reached; at that moment, part of the
fat was crystallized and part was not. The rate of crystallization,
which mainly depends on the heat release from the sample to the
environment, was different in the textural and ultrasonic measure-
ments due to the experimental set-ups used. In the case of the
textural measurements, the samples were placed in a Petri dish
in contact with air, whereas a metallic cylinder in contact with
stirred water was used for the ultrasonic measurements. Thereby,
the heat transfer was much faster in the velocity measurements
making that the beginning of the steep increases (due to the crys-
tallization of the different TGs fractions) took place earlier in the
velocity experiments than in the textural ones. In the experi-
ments at 10 C, only the rst increase in the ultrasonic velocity
was observed, which then remained constant until the end of the
storage (480 h). At 10 C, the most unsaturated TGs did not
crystallize and, consequently, only the rst increase in the ultra-
sonic velocity was observed. This behavior also occurred during
the textural measurements at the highest cold storage temperature
(Figure 3).
Ultrasonic measurements showed that the higher the storage
temperature, the lower the ultrasonic velocity. In the case of
Montanera fat, the ultrasonic velocity increased from 1525 to
1705 m/s at 10 C and 0 C respectively (Figure 4A), whereas the
ultrasonic velocity in Cebo changed from 1570 to 1800 m/s for
the same range. As previously mentioned, high crystallization tem-
peratures lead to larger fat crystals and a softer crystalline structure
than low temperatures. In addition, it should be remarked that the
ultrasonic velocity in liquid and solid fats also falls as the temper-
ature rises (Benedito and others 2001).
E834 Journal of Food Science
r
Vol. 79, Nr. 5, 2014
E
:
F
o
o
d
E
n
g
i
n
e
e
r
i
n
g
&
P
h
y
s
i
c
a
l
P
r
o
p
e
r
t
i
e
s
Ultrasonics monitoring crystallization . . .
Finally, in a similar way to textural measurement, the ultra-
sonic velocity allowed the 2 types of fats to be differentiated.
When comparing the curves of Montanera and Cebo fats (Fig-
ure 4), the highest values of ultrasonic velocity at the same storage
temperature were found in Cebo. Thus, at 0 C, the maximum
ultrasonic velocity was 90 m/s higher for Cebo than Montan-
era (1800 versus 1710 m/s). As explained in section 3.3, the
higher level of saturation in Cebo TGs implies a harder crys-
talline structure in which ultrasound is propagated faster than in
Montanera.
Therefore, ultrasound permitted the characterization of the
change in the textural properties brought about by crystalliza-
tion during cold storage. Furthermore, the crystallization pattern
will depend on the TGs composition, which is characteristic for
each breed and rearing system (Petr on and others 2004; Ni noles
and others 2007).
The state of the fat has a great effect on the quality of the dry-
cured meat products (texture and aroma release) which contain
Iberian fat. Texture of meat is affected by the fat amount and state
because it has a different resistance to shearing and an easy frac-
ture compared to other muscle components. Therefore, ultrasound
could also be used as an alternative to DSC or textural methods to
identify the overall quality of pork-meat products and to monitor
the changes that take place during cold storage. This application
is highly relevant as ultrasound could be measured on the surface
of vacuum packaged products like sliced dry-cured ham (pack-
age thickness lower than 2.5 cm) and used as a non-destructive
technology which is easily automated online. The effect of the
packaging on the ultrasonic measurements should be taken into
account, also the effect of the mixture of fat and lean tissue on the
ultrasonic velocity should be considered.
Modeling of the change in the ultrasonic velocity and
textural properties during cold storage
Modeling the ultrasonic velocity change is relevant for predict-
ing and quantifying the crystallization process during the cold stor-
age of pork-meat products. A mathematical model developed by
Santacatalina and others (2010), based on the Avrami model, was
applied to describe the evolution of the ultrasonic velocity during
crystallization of both Montanera and Cebo fats (Figure 4).
Table 3 shows the results obtained for the modeling of the ul-
trasonic velocity. The Avrami model tted the experimental data
of ultrasonic velocity well. In Figure 4, it may be observed that
the model followed the same trend as the experimental data, the
percentage VAR and RMSE values being higher than 99.5% and
lower than 3.5 m/s for Cebo fat and higher than 99.7% and lower
than 2.3 m/s for Montanera fat, respectively. Both statistical pa-
rameters indicated that the 2-step crystallization model developed
1500
1550
1600
1650
1700
1750
1800
1850
0 120 240 360 480
V

(
m
/
s
)
Time (h)
0 C
2 C
5 C
7 C
10 C
Model
1500
1550
1600
1650
1700
1750
1800
1850
0 120 240 360 480
V

(
m
/
s
)
Time (h)
0 C
2 C
5 C
7 C
10 C
Model
A
B
Figure 4Changes and modeling of the
ultrasonic velocity during cold storage at
different temperatures of Montanera (A)
and Cebo (B) fats. Average values
Standard deviation are plotted from
3 replicates.
Vol. 79, Nr. 5, 2014
r
Journal of Food Science E835
E
:
F
o
o
d
E
n
g
i
n
e
e
r
i
n
g
&
P
h
y
s
i
c
a
l
P
r
o
p
e
r
t
i
e
s
Ultrasonics monitoring crystallization . . .
Table 3Results for the modeling of the ultrasonic velocity during cold storage of Montanera and Cebo fats at different tem-
peratures (0 C, 2 C, 5 C, and 7 C). Model parameters, percentage of explained variance (VAR) and root mean square error
(RMSE).
T a
1
K
1
a
2
K
2
VAR RMSE
(C) (m/s) (s
n
) n
1
(m/s) (s
n
) n
2
(h) (s) (%) (m/s)
Montanera
0 1647 6.9 10
3
0.4 1708 3.7 10
06
1.2 7.4 7.4 10
3
99.8 1.8
2 1587 7.9 10
4
0.8 1689 2.2 10
08
1.5 23.3 33.7 10
3
99.9 0.8
5 1564 1.5 10
4
0.9 1673 3.0 10
11
1.9 134.0 441 10
3
99.7 2.3
7 1555 3.9 10
3
0.5 1602 7.5 10
09
1.9 257.5 435 10
3
99.8 1.6
10 1527 2.0 10
1
0.1 81.5 0.6
Cebo
0 1666 3.0 10
3
0.7 1808 1.8 10
2
0.4 3.9 14.0 10
3
99.9 0.9
2 1636 5.0 10
3
0.6 1753 4.7 10
6
2.2 8.4 16.3 10
3
99.6 2.8
5 1621 3.0 10
3
0.6 1742 1.9 10
3
2.0 42.1 142 10
3
99.5 3.5
7 1606 5.8 10
2
0.3 1714 7.0 10
3
3.0 167.3 1556 10
3
99.7 2.2
10 1572 1.3 10
2
0.5 77.4 0.9
Subscripts 1 and 2 denote the rst and second steep increases in the ultrasonic velocity versus time curve, respectively.
by Santacatalina and others (2010) was appropriate for modeling
the change which took place in the ultrasonic velocity during the
cold storage of the 2 types of fat used in this work.
The crystallization temperature affected the a
1
and a
2
parame-
ters. These parameters are the values of the plateau in the rst and
second increases in the ultrasonic velocity when the crystallization
of both saturated and unsaturated TGs is completed. Thus, for
both parameters, the lower the temperature, the higher the ve-
locity plateau. As already mentioned, the faster the crystallization
(low temperatures), the harder the sample and, as a consequence,
the higher the velocity. The relationship between a
1
and a
2
and
the crystallization temperature was well described by signicant
(P < 0.05) linear regressions. For the 0 C to 10 C temperature
range, a linear relationship (Cebo Eq. 5 and 6 and Montanera
Eq. 7 and 8) was found between the parameters a
1
and a
2
and the
crystallization temperature:
a
1
= 8.7(T) +1661 R
2
= 0.97 (5)
a
2
= 11.8(T) +1795 R
2
= 0.88 (6)
a
1
= 10.7(T) +1628 R
2
= 0.89 (7)
a
2
= 13.6(T) +1716 R
2
= 0.84 (8)
where T is the crystallization temperature (C).
These equations could be used to estimate the maximum ve-
locity and, therefore, the maximum texture reached by a sample
stored in a temperature range of 0 C to 10 C. As an example,
if Cebo fat was crystallized at 6 C, the a
1
and a
2
values would
be 1609 and 1724 m/s, respectively. On the other hand, when
comparing Montanera and Cebo fats (Table 3), the highest values
of the a
1
and a
2
parameters for each temperature were obtained for
Cebo. As already mentioned, the higher content of saturated fatty
acids present in Cebo fat leads to a more compact TGs packaging,
involving higher values of the elastic modulus, which, in turn,
leads to a faster propagation of the ultrasonic wave. Therefore, the
parameters a
1
and a
2
allow the different crystallization pattern of
Montanera and Cebo fats to be identied and quantied.
The temperature was also found to inuence the parameter of
the modied Avrami equation. This parameter represents the time
(h) when the second crystallization stage starts and is characteristic
of the onset time of crystallization of the most unsaturated TGs.
In this case, exponential equations were identied to quantify the
inuence of the crystallization temperature on the parameter
(Cebo Eq. 9 and Montanera Eq. 10):
= 3.35 e xp (0.537 T) R
2
= 0.99 (9)
= 8.03 e xp (0.519 T) R
2
= 0.99 (10)
Using Eq. 9, therefore, it could be predicted that, at 6 C, the
crystallization of the most unsaturated TGs in Cebo would begin at
84 h. At high temperatures, the exponential relationship indicates
that the crystallization of the most unsaturated TGs would only
occur at very long storage times. Thus, although Eq. 9 and 10
are only valid in the range of 0 C to 7 C, the parameter at
10 C for Montanera should be 1441 h, which could explain why
the crystallization of the most unsaturated TGs was not observed
for 5 storage days (480 h) at this temperature (Figure 4). However,
it could also be possible that the most unsaturated TGs cannot
crystallize at 10 C. The main inuence of the origin of the fat
on the relationship between and temperature was found in the
pre-exponential factor, which was 3.35 in the case of Cebo and
8.03 for Montanera (Eq. 9 and 10), respectively. This means that
the crystallization of the most unsaturated TGs in Montanera is
delayed if compared to that in Cebo, which can also be seen in
Figure 4.
As reported in section 3.4, similar behavior between the ul-
trasonic velocity (Figure 4) and the textural changes (Figure 3)
was observed during cold storage. Therefore, the Avrami equa-
tion was also applied to describe the changes in the MF, related
with the hardness of the samples (Figure 3) during cold storage.
In this regard, the changes in the penetration force as a function
of crystallization time were modeled using Eq. 11:
FM(t ) = FM
0
+a

(1 exp(k

t
n

)) (11)
where FM is the maximum penetration force (N) at time t (s), FM
0
the maximum penetration force at the beginning of the isothermal
crystallization, a the maximum force at innite time, k is the
crystallization rate constant (s
n
) dependent on the crystallization
temperature and n is the Avrami exponent (dimensionless).
E836 Journal of Food Science
r
Vol. 79, Nr. 5, 2014
E
:
F
o
o
d
E
n
g
i
n
e
e
r
i
n
g
&
P
h
y
s
i
c
a
l
P
r
o
p
e
r
t
i
e
s
Ultrasonics monitoring crystallization . . .
Table 4Results for the modeling of the maximum penetration
force during cold storage of Montanera and Cebo fats at dif-
ferent temperatures (0 C, 2 C, and 5 C). Model parameters,
percentage of explained variance (VAR) and root mean square
error (RMSE).
T a
1
K
1
a
2
K
2
VAR RMSE
(C) (N) (s
n
) n
1
(N) (s
n
) n
2
(h) (s) (%) (N)
Montanera
0 2.7 7.6 10
04
0.8 8.1 8.1 8.2 9.0 33.2 10
3
95.3 0.09
2 2.1 1.5 10
06
1.3 6.5 8.0 12.2 17.8 69.1 10
3
95.4 0.05
5 2.1 3.5 10
07
1.2 66.9 0.02
Cebo
0 9 1.7 10
04
1.0 17.0 3.6 10
08
1.6 1.0 65.4 10
3
97 0.13
2 5.2 1.0 10
04
0.9 11.0 8.05 7.8 32.8 43.9 10
3
94.3 0.09
5 5.2 3.3 10
02
1.2 58.5 0.04
Subscripts 1 and 2 denote the rst and second steep increases in the maximum
penetration force versus time curve, respectively.
Equation 11 was modied to include the 2-step crystallization
process (Eq. 12):
FM = A

1
Avr

1
+ A

2
Avr

2
(12)
where Avr
1
and Avr
2
are the Avrami equations (Eq. 11) for the
rst and second maximum force increases, respectively, and A
1

and A
2
are calculated in a similar way to the ultrasonic velocity
using Eq. 3 and 4.
Figure 3 shows the t of the Avrami model for both Cebo and
Montanera fats; in spite of the great variability of the experimental
data, the model adequately described the behavior of the textural
changes during cold storage. The results showed the inuence of
the crystallization temperature on a
1
and a
2
parameters (Table 4);
therefore, the lower the temperature used, the higher the max-
imum force and the harder the sample. When comparing the 2
types of fat at the same temperature (0 C; Table 4), the highest
values of the a
1
and a
2
parameters were obtained for Cebo (9.0
and 17.0 N) and the lowest for Montanera (2.7 and 8.1 N). As
already explained, this fact arises from the differing fatty acid pro-
le that leads to a different TGs crystallization. The differences in
the connective tissue (Ni noles and others 2008) could also lead to
the variations in the texture of the 2 types of fat. Therefore, the
modied Avrami model was adequate for the description, quan-
tication and prediction of the ultrasonic and textural changes that
took place during cold storage as a result of the fat crystallization.
In addition, the model could be used to differentiate between fats
with different composition.
Conclusions
In this work, textural changes were observed in 2 types of
backfats (Montanera and Cebo Iberian fats) during cold storage.
Fat hardness, instrumentally characterized as MF, showed 2 steep
increases due to the crystallization of the most saturated and un-
saturated TGs. During crystallization, Cebo fat was always harder
than Montanera, which was explained by taking their different
fatty acid composition into account. Ultrasonic velocity permit-
ted an optimum characterization of textural changes during fat
crystallization. Therefore, this technology successfully identied
the TGs crystallization and characterized the different crystalliza-
tion pattern of fats with different composition. Finally, a 2-step
model based on the Avramis equation was used to describe the
textural and ultrasonic changes in both Montanera and Cebo fats
during the crystallization process.
Therefore, the results obtained in this work indicate that the
ultrasonic technique could be a reliable method with which to
characterize the crystallization pattern of fats during cold stor-
age, which largely affect the quality of dry-cured meat products.
Further work should be carried out in the future to establish the
feasibility of using non-destructive ultrasonic techniques in the be-
havior characterization of dry-cured products during cold storage,
to identify the progress of crystallization and the origin of the fat.
Acknowledgments
The authors acknowledge the nancial support from the
Ministerio de Economa y Competitividad in Spain (Project
RTA 2010-00029-C04-02).
Author Contributions
Edith Corona conducted the experiments, analyzed the data and
drafted the document. Juan V. Santacatalina contributed in the ul-
trasonic measurements and data analysis. Sonia Ventanas conducted
the chemical analyses. Jos e V. Garca-P erez and Jos e Benedito de-
signed the study, supervised the data analysis and reviewed the
manuscript.
References
Benedito J, Carcel JA, Gonz alez R, SanJuan N. 2000. Prediction of Instrumental and Sensory
textural characteristics of Mahon cheese fromultrasonic measurements. J Texture Stud 31:631
43.
Benedito J, Carcel JA, Rosell o C, Mulet A. 2001. Composition assessment of raw meat mixtures
using ultrasonics. Meat Sci 57:36570.
BOE 2007. Real Decreto 1469/2001 por el que se aprueba la norma de calidad para el jam on
Ib erico, paleta ib erica y ca na de lomo ib erico elaborados en Espa na. Boletn Ocial del Estado
de 3 de Noviembre de 2007.
Campos R, Narine SS, Marangoni AG. 2002. Effect of cooling rate on the structure and
mechanical properties of milk fat and lard. Food Res Int 35:97181.
Carrapiso AI, Bonilla F, Garca C. 2003. Effect of crossbreeding and rearing system on sensory
characteristics of Iberian ham. Meat Sci 65:6239.
Cebula DJ, Smith KW. 1991. Differential scanning calorimetry of confectionery fats. Pure
tryglicerides: effects of cooling and heating rate variation. J Am Oil Chem Soc 68:591
5.
Cebula DJ, Smith KW. 1992. Differential scanning calorimetry of confectionery fats. Part II:
effects of blends and minor components. J Am Oil Chem Soc 69:9928.
Cerdeira M, Candal RJ, Herrera ML. 2004. Analytical techniques for nucleation studies in lipids:
advantages and disadvantages. J Food Sci 69:R18591.
Cilla I, Martnez L, Beltr an JA, Roncal es P. 2006. Effect of low-temperature preservation on
the quality of vacuum-packaged dry-cured ham: refrigerated boneless ham and frozen ham
cuts. Meat Sci 73:1221.
Corona E, Garca-P erez JV, Mulet A, Benedito, J. 2013. Ultrasonic assessment of textural changes
in vacuum packaged sliced Iberian ham induced by high pressure treatment or cold storage.
Meat Sci 95:38995.
Folch J, Lees M, Sloane-Stanley GH. 1957. A simple method for the isolation and purication
of total lipids from animal tissues. J Biol Chem 226:497509.
Foubert I, Vanrolleghem PA, Thas O, Dewettinck K. 2004. Inuence of chemical composition
on the isothermal cocoa butter crystallization. J Food Sci 69:E47887.
Fuentes V, Ventanas J, Est evez M, Carrapiso AI, Ventanas S. 2010. Application of time-intensity
technique to study the effect of fat composition on texture and avour perception in a dry-
cured model product. Oral presentation in the Fourth European Conference on Sensory and
Consumer Research (Vitoria, Spain).
Garca-Esteban M, Ansorena D, Astiasar an I. 2004. Comparison of modied atmosphere pack-
aging and vacuum packaging for long period storage of dry-cured ham: effects on color,
texture and microbiological quality. Meat Sci 67:5763.
Himawan C, Starov VM, Stapley AGF. 2006. Thermodynamic and kinetic aspects of fat crys-
tallization. Adv Colloid Interface Sci 122:333.
Kalnin D, Lesieur P, Artzner F, Keller G, Ollivon M. 2005. Systematic investigation of lard
polymorphism using combined DSC and time-resolved synchrotron X-ray diffraction. Eur J.
Lipid Sci. Technol 107:594606.
Le Mestre M, Cornily G, Simatos D. 1984. Le comportement termique des lipides musculaires
et de d ep ot chez le porc. Revue Francaise des Corps Gras 31:10715.
Llull P, Simal S, Femenia A, Benedito J, Rossell o C. 2002. The use of ultrasound velocity
measurement to evaluate the textural properties of sobrassada from Mallorca. J Food Eng
52:32330.
Maleky F, Campos R, Marangoni AG. 2007 Structural and mechanical properties of fats quan-
tied by ultrasonics. J Am Oil Chem Soc 84:3318.
Martini S. 2013. Sonocrystallization of Fats, SpringerBriefs in Food, Health, and Nutrition.
New York: Springer.
Martini S, Bertoli C, Herrera ML, Neesond I, Marangoni A. 2005a. Attenuation of ultrasonic
waves: inuence of microstructure and solid fat content. J Am Oil Chem Soc 82:31928.
Vol. 79, Nr. 5, 2014
r
Journal of Food Science E837
E
:
F
o
o
d
E
n
g
i
n
e
e
r
i
n
g
&
P
h
y
s
i
c
a
l
P
r
o
p
e
r
t
i
e
s
Ultrasonics monitoring crystallization . . .
Martini S, Bertoli C, Herrera ML, Neesond I, Marangoni A. 2005b. In situ monitoring of solid
fat content by means of pulsed nuclear magnetic resonance spectrometry and ultrasonics. J
Am Oil Chem Soc 82:30512.
Mizrach A, Flitsanov U. 1999. Nondestructive ultrasonic determination of avocado softening
process. J Food Eng 40:13944.
Mulet A, Benedito J, Bon J, Sanju an N. 1999. Low intensity ultrasonics in food technology.
Food Sci Technol Int 5:28597.
Nielsen M, Martens HJ. 1997. Low frequency ultrasonics for texture measurements in cooked
carrots. Trans ASAE 40:17418.
Ni noles L, Clemente G, Ventanas S, Benedito J. 2007. Quality assessment of Iberian pigs
through backfat ultrasound characterization and fatty acid composition. Meat Sci 76:102
11.
Ni noles L, Sanjuan N, Ventanas S, Benedito J. 2008. Ultrasonic and sensory characterization of
dry-cured ham fat from Iberian pigs with different genetics and feeding backgrounds. Meat
Sci 80:896902.
Petr on MJ, Muriel E, Tim on ML, Martn L, Antequera T. 2004. Fatty acids and triacylglycerols
proles from different types of Iberian dry-cured hams. Meat Sci 68:717.
Rubio B, Martnez B, Garca-Cach an MD, Rovira J, Jaime I. 2007. The effects of high pressure
treatment and of storage periods on the quality of vacuum-packed salchich on made of raw
material enriched in monounsaturated and polyunsaturated fatty acids. Innov Food Sci Emerg
Technol 8:1807.
Ruiz-Carrascal J, Ventanas J, Cava R, Andr es AI, Garca C. 2000. Texture and appearance of
dry cured ham as affected by fat content and fatty acid composition. Food Res Int 33:915.
Sandler SR, Karo W. 1992. Source book of advances organic laboratory preparations. San Diego
Calif.: Academic Press.
Santacatalina JV, Garcia-Perez JV, Corona E, Benedito J. 2010. Ultrasonic monitoring of lard
crystallization during storage. Food Res Int 44:14655.
Singh AP, McClements DJ, Marangoni AG. 2004. Solid fat content determination by ultrasonic
velocimetry. Food Res Int 37:54555.
Tim on ML, Martn L, Petr on MJ, Jurado A, Garcia C. 2001a. Composition of subcutaneous
fat from dry-cured Iberian hams as inuenced by pig feeding. J Sci Food Agr 82:186
91.
Tim on ML, Ventanas J, Carrapiso AI, Jurado A, Garca C. 2001b. Subcutaneous and intermus-
cular fat characterization of dry-cured Iberian hams. Meat Sci 58:8591.
Ventanas S, Est evez M, Andr es AI, Ruiz J. 2008a. Analysis of volatile compounds of Iberian dry-
cured loins with different intramuscular fat contents using SPMEDED. Meat Sci 79:172
80.
Ventanas S, Tejeda FJ, Est evez M. 2008b. Chemical composition and oxidative status of tissues
from Iberian pigs as affected by diets: extensive feeding v.oleic acid and tocopherol-enriched
mixed diet. Animal 2:62130.
Ventanas S, Ventanas J, Ruiz J, Est evez M. 2005. Iberian pigs for the development of high-quality
cured products. Res Dev Agric Food Chem 6:2754.
Yucel U, Coupland JN. 2011. Ultrasonic characterization of lactose crystallization in gelatin gels.
J Food Sci 76:E4854.
E838 Journal of Food Science
r
Vol. 79, Nr. 5, 2014