Journal of Food Engineering 107 (2011) 289295

Contents lists available at ScienceDirect

Journal of Food Engineering

journal homepage: www.elsevier.com/locate/jfoodeng

Prediction of out-of-container pasteurization of pickled cucumbers using the

nite-difference method
Rubn O. Morawicki a,, Miguel E. Schmalko b
a
b

Department of Food Science, University of Arkansas, 2650 North Young Ave., Fayetteville, AR 72703, United States
Facultad de Ciencias Exactas, Qumicas y Naturales, Universidad Nacional de Misiones, Flix de Azara 1552, 3300 Posadas, Misiones, Argentina

a r t i c l e

i n f o

Article history:
Received 21 March 2011
Received in revised form 7 July 2011
Accepted 15 July 2011
Available online 23 July 2011
Keywords:
Fermented cucumber
Pasteurization
Finite-difference method
High-acid food
Peroxidase

a b s t r a c t
In-container pasteurization of pickled cucumbers deactivates enzymes and spoilage microorganisms
with minimal organoleptic changes when packed in small jars. As jar volumes increase, pickle quality
is affected as a result of longer pasteurization times setting the practical volume limit to 3.8 L. Alternatively, this paper proposes out-of-container pasteurization followed by packaging in sanitized containers.
This study, including only the pasteurization step, was conducted with a discontinuous pasteurization
device using variable cucumber diameters, brine temperatures, and ows. Heat penetration data was
t to a mathematical model with the enzyme peroxidase as a parameter of pasteurization effectiveness.
Results showed that heat transfer was dominated by internal resistance, and according to the model,
enzyme deactivation depended on the cucumber diameter, and brine temperature, but not on the brine
ow. A 5-log reduction in peroxidase activity in the center of the cucumbers required 1.4 times longer
than the equivalent treatment in the whole pickle.
2011 Elsevier Ltd. All rights reserved.

1. Introduction
In-container pasteurization is the nal step in the production of
shelf-stable pickled cucumbers to extend their shelf-life. Heat pasteurization not only is effective in reducing the number of spoilage
microorganisms, but also in deactivating enzymes that participate
in quality deterioration, mainly due to tissue softening. Pasteurization, however, alters the organoleptic quality of pickles, so the
intensity of the treatment (time and temperature) must be minimized. Since pickled cucumbers have a pH lower that 4.6, they
are considered an acid food with no risk of Clostridium botulinum
growth. Accordingly, the pasteurization parameters currently used
by the pickling industry are used only for slowing the deterioration
caused by microorganisms and enzymes. The most common heat
treatment for pasteurization of pickled cucumbers in industry is
74 C (165F) for 15 min in the center of a cucumber located in
the center of the jar (Breidt et al., 2005).
As with any other in-container heat pasteurization or sterilization, the total heating duration to achieve the desired pasteurization units, or lethality, in the center of the jar depends on various
factors including container size. Since the external heat transfer
area of jars does not increase proportionally with the volume, the
length of heat treatments does not increase proportionally with
volumes. For instance, as the volume of the jar triples, the external
Corresponding author. Tel.: +1 479 575 4923; fax: +1 479 575 6936.
E-mail address: rmorawic@mail.uark.edu (R.O. Morawicki).
0260-8774/$ - see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jfoodeng.2011.07.020

area increases approximately doubles. Therefore, larger jars need

longer pasteurization times and pickles located in the periphery
of the container are exposed to more heat than the ones in the center. This problem imposes a limit to the maximum volume that can
be subjected to in-container pasteurization to approximately 1 gal
(3.8 L); and containers larger than 1 gal, which include 2- and 5-gal
institutional pails, are not pasteurized.
To eliminate the effect of the container and to have greater control of pasteurization, this paper proposes a different approach
similar to aseptic processing. In aseptic processing, food sterilization takes place outside the containers, and the sterile food is
packed in sterile containers in a sterile environment. In a similar
manner, pickles could be pasteurized prior to packaging and then
packaged in sanitized containers inside a clean environment.
To study out-of-container pasteurization, pickled cucumbers of
three different diameters were subjected to pasteurization with
hot brine at three ow rates and three different temperatures
while the internal temperature of the cucumbers was recorded.
The resulting heating curves were used to validate a mathematical
model that described the degree of pasteurization as a function of
time using the nite-difference method. The nite-difference
method is considered a suitable approach to predict internal temperature prole of various food products (Lin et al., 2009; Martens
et al., 2001; Mattos et al., 2005; Wang et al., 2007). For this research, the model was built by dividing a hypothetical cucumber
in concentric layers (nodes), and it was assumed that the temperature remained constant in each node during a nite time step.

290

R.O. Morawicki, M.E. Schmalko / Journal of Food Engineering 107 (2011) 289295

Nomenclature
Ac (k)
As (k)
Cp
D
Dc
Ds
DRc
DRs
hc
hs
kw
kT

cylinder area between nodes k-1, and k (m2)

sphere area between nodes k-1, and k (m2)
specic heat capacity (kJ/kgK)
decimal reduction time (min1)
cylinder diameter of the cucumber (M)
semi sphere diameter of the cucumber (M)
cylinder node thickness (M)
sphere node thickness (M)
external heat transfer coefcient of the cylinder
(J/m2Ks)
external heat transfer coefcient of the sphere (J/m2Ks)
thermal conductivity of the water (W/mK)
thermal conductivity of the cucumber (W/mK)

Then with the experimental temperature proles, the degree of the

enzyme degradation was estimated by taking into account the contribution of each node at each time interval.
The specic aim of this research was to investigate the effect of
brine ow, brine temperature, and cucumber diameter on attaining a 5-logarithmic cycle reduction of the enzyme peroxidase
during out-of-container pasteurization of pickled cucumbers.
Experimental data was used to validate a model developed with
the nite-difference method considering only the heating phase
of pasteurization and the reduction of peroxidase activity in both
the center of the cucumber and the average reduction in the whole
cucumber.
2. Materials and methods
2.1. Determination D-value and z-value of peroxidase enzyme
2.1.1. Materials
Fresh pickling cucumber fruits (Cucumis stativus) of an unidentied variety were obtained from a local market and used for the
determination of the D-value and z-value for peroxidase
deactivation.
2.1.2. Preparation of cucumber extracts for peroxidase determination
The procedure for peroxidase determination was adapted from
Buescher and McGuire (1986). All glassware, solutions, and materials were cooled to 4 C prior to use. Refrigerated cucumbers were
diced with a sharp blade into small cubes and, after mixing the
dices thoroughly, a 50-g aliquot was taken and combined with
100 mL of 1.0 M NaCl solution in citratephosphate buffer (pH
6.5). The mixture was blended in a previously cooled Waring blender Model HGB7WTG4 (Torrington, CT) at speed 7 for 1 min. The
crude homogenate was immediately transferred to an Erlenmeyer
ask, refrigerated for 1 h at 4 C, and then ltered through four layers of cheesecloth. The ltrate was immediately refrigerated and
the solids resuspended in 50 mL of 1.0 M NaCl in citrate buffer
and blended at speed 7 for 1 min. The second crude homogenate
was ltered through cheesecloth and the resulting ltrate mixed
with the initial ltrate. The liquid was then centrifuged at 4 C
and 10,000g for 10 min and immediately used for the determination of the D-value of peroxidase deactivation.
2.1.3. Analysis of peroxidase activity
Peroxidase activity was determined by following the degradation of the chromogen 3,30 ,50 ,5-Tetramethylbenzidine (TMB) obtained from SigmaAldrich Co. (St. Louis, MO). TMB (1.45 mL), at
25 C, was transferred with a pipette to a spectrophotometer cuvette followed by the addition of 0.05 mL of the cucumber extract.

MPE
Pr
Re
RMSE
Tc (k,t)
Tm
Ts(k,t)
Vc(k)
Vs(k)
z
Dt

mean percent error

Prandt number of water
Reynold number of water
root of the mean square error (C)
temperature of the cylinder node k, at time t (C)
medium temperature (C)
temperature of the sphere node k, at time t (C)
cylinder node volume (m3)
sphere node volume (m3)
thermal resistant constant (C)
time increment (s)
cucumber apparent density (kg/m3)

Immediately after mixing, the absorbance at 370 nm was determined by a spectrophotometer (Beckman DU-520, Fullerton, CA)
and used as an indicator of the peroxidase activity (Sigma, 1995).
2.1.4. Determination of D- and z-value
The initial peroxidase activity of the cucumber extract was
determined and used as the maximum peroxidase activity. Then
aliquots of the extract (1 mL) were transferred to 2-mL Eppendorf
tubes and incubated in water baths set at 71.1, 73.9, 76.7, and
79.4 C (160, 165, 170, and 175F, respectively). Periodic samples
were taken and assayed for peroxidase activity as follows: every
15 min for extracts held at 71.1 C, every 10 min for extracts held
at 73.9 and 76.7 C, and every 5 min for extracts held at 79.4 C.
Incubation was continued until the peroxidase activity was reduced 2-logarithmic cycles or more compared to the starting activity. For each temperature, assays were conducted in triplicate on
different days and with different lots of cucumbers. The peroxidase
activity was then standardized by dividing the activity at every
sampling point by the initial activity. Standardized triplicates were
averaged and plotted in an xy scatter plot with a logarithmic scale
for enzyme activity and a regular scale for time. The data points
were then t with a linear regression line using the least-squares
criterion. D-value was calculated as the time required reducing
the peroxidase activity 1-logarithmic cycle.
The z-value for peroxidase deactivation was obtained by plotting D-values for each temperature in an xy scatter plot with a
logarithmic scale for the D-value and a regular scale for the temperature. A regression line was established as described for the
D-value determination) and the z-value was calculated as the increase in temperature required reducing the D-value by one logarithmic cycle.
2.2. Heat penetration during pasteurization
2.2.1. Materials
Commercial pickled cucumbers, Vlasic Kosher Dills packed in
1.3 and 2.36-L jars (Pinnacle Foods Corporation, Cherry Hill, NJ),
Gedney Dill Whole Pickles packed in 3.78-L jars (M.A. Gedney
Co., Chaska, MN), and Mt. Olive Kosher Dills packed 3.78-L jars
(Mt. Olive Pickle Co. Inc., Mt. Olive, NC), were purchased from a local grocer. From these jars, straight, undamaged pickles with
lengths between 0.08 and 0.125 m were selected, from which three
diameters were chosen (2.5  102 m, 3.3  102 m, and
4.0  102 m) and used for the pasteurization experiments.
2.2.2. Pasteurization device
The apparatus consisted of: (i) a perforated tray of 0.25  0.3 m
that held the cucumbers; (ii) a brine-collecting tray located below

R.O. Morawicki, M.E. Schmalko / Journal of Food Engineering 107 (2011) 289295

Brine
dispenser
Cucumber

291

from the brine with a thin layer of lacquer. Thermocouples were

connected to a portable data logger, OM-DAQPRO-5300 (Omega
Engineering, Inc., Stamford, CT), that was set to acquire temperatures every 15 s for a total of 1500 s. Data sets were then transferred to a personal computer with the DaqPro software for the
OM-DAQPRO-5300 data. After each experimental run, cucumbers
were cut and the distances between the thermocouples and the
center measured.

4
2.3. Mathematical model

Flow
meter
3
2

Booster
pump

Heater/Pump

Brine
reservoir

1; 2; 4; and 5: ball valves

3: needle valve
Fig. 1. Pasteurization device.

the perforated tray; (iii) a 20-liter brine tank; (iv) a brine dispenser
on top of the cucumbers; (v) a heater/pump; (vi) a booster pump;
(vii) a volumetric variable area ow meter; and (viii) stainless steel
manually operated ball valves connected with silicone tubing
(Fig. 1). The brine dispenser was a 0.13-m-diameter perforated disk
with 20 holes (each with a diameter of 0.03 m) equally distributed
throughout the disk that dispensed the brine as a shower. The heater/pump combination was a VWR immersion circulator, model
1122S (VWR, West Chester, PA), that was used to heat the brine
to the desired temperature. The variable area ow meter was an
F-440 (Blue-White Industries, Ltd., Huntington Beach, CA) with a
range between 1.5 and 18 L/min (LPM). The booster pump was a
Wayne Reliant One Model 75982 (Wayne Water Systems, Harrison
OH).
2.2.3. Experimental design
The experimental design was a full factorial with three factors
(cucumber diameter, brine temperature, and brine ow) at three
levels. The levels were 2.5  102 m, 3.3  102 m, and
4.0  102 m for cucumber diameter, 76.7, 79.4, and 82.2 C (170,
175 and 180F) for brine temperature, and 6.4, 12.8, and
19.2  105 m3/s (1, 2, and 3 gal/min) for brine ow.
Each experimental run used three cucumbers that were placed
on the sample holding tray (as shown in Fig. 1) with two of them
connected via thermocouples to the data acquisition system.

The model was developed according to the nite-difference

method. For modeling, the cucumber was assumed to be a cylinder
with two hemispheres at the ends. Both, the cylinder and the
hemispheres, were divided into 15 concentric nodes of variable
thickness between 1.67  103 and 2.67  103 m that varied
in proportion to the cucumber diameters (Fig. 2). It was assumed
that nodes in the cylindrical section continued to the semispherical
zone, and this assumption was validated by comparing temperatures of adjacent nodes in the cylinder and the hemispheres. Based
on a comparison with unsteady-state heat gures for a cylinder
and a sphere (Fig. 5.11 at Perry and Green, 1997), it was concluded
that differences were insignicant. This indicated the absence of a
longitudinal temperature prole; therefore, a model considering
unidirectional heat ow was developed for each section.
In each node, an energy balance in unsteady-state was conducted during a Dt time increment (Hoffman, 1992). Using the energy balance, the temperature after Dt increments was calculated
for each node. The Dt used for the model was 0.5 s, and it was assumed that the node temperature remained constant during this
period. The resulting equations for the model were as follows:
For the cylinder:
Node k = 1 (external node)


Dt
 hc  Ac0  Tm  Tc1; t
q  Cp  Vc1

kT  Ac1
 Tc1; t  Tc2; t
1

DRc

Tc1; t 1 Tc1; t

Nodes k = 2 to n  1

Dt
TcK; t 1 TcK; t
q  Cp  VcK
"
#
 Tck  1; t  Tck; t
c kT Ack1
DRc

 Tck; t  Tck 1; t
 kT AcK
DRc

Node k = n = 15 (at the center)

Dt
TcN; t 1 TcN; t
q  Cp  VcN


kT  AcN  1

 TcN  1; t  TcN; t
DRc

For the semi spheres:Node k = 1 (external node)


Dt
 hs  As0  Tm  Ts1; t
q  Cp  Vs1

kT  As1
 Ts1; t  Ts2; t
4

DRs

Ts1; t 1 Ts1; t

Nodes k = 2 to n  1
2.2.4. Temperature data acquisition
Temperature during heat treatments was acquired by two thermocouples located inside the cucumbers. One was placed as close
to the center as possible and the second one at the periphery just
under the skin. The thermocouples were Type T, model number
39138-T manufactured by Atkins (Cooper-Atkins Corporation,
Gainesville, FL), with bare tips, which were electrically insulated

Dt
Tsk; t 1 Tsk; t
q  Cp  VsK
"
#
 Tsk  1; t  Tsk; t
c kT Ask1
DRs

 Tsk; t  Tsk 1; t
 kT Ask
DRs
Node k = n = 15 (at the center)

292

R.O. Morawicki, M.E. Schmalko / Journal of Food Engineering 107 (2011) 289295

Fig. 2. Cucumber adapted to the shape of a cylinder with hemispherical ends divided in zones and nodes.

(a)

(b)
Fig. 3. Decrease in the activity of peroxidase extracted from fresh cucumbers and
incubated at four different temperatures for various durations.

Dt
Tsn; t 1 Tsn; t
q  Cp  VsN


kT  Asn  1

 Tsn  1; t  Tsn; t
DRs

Simulations with Matlab V 7.4.0.287 (R 2007a) (The Math

Works Inc., MA, USA) predicted the temperature proles using
the model. The tness of the model was validated by comparing
the predicted and experimental values every 15 s using the mean
percentage error (MPE) (Eq. (7)) and the root mean square error
(RMSE) (Eq. (8)) as indicators. The MPE was calculated from predicted and experimental temperatures.

MPE

n
100 X
ABST exp  T pre
n 1
T exp

s
n
1X
T exp  T pre 2
RMSE
n 1

The degree of heat treatment, indicated by the activity reduction of the enzyme peroxidase, was estimated by adding the heat
treatment in each node during a Dt interval (Eq. (9)):

log

n
C0 X
Dt
Vnode


D  v alue Vtotal
C
1

(c)

The overall treatment was calculated by adding the values of C0/

C for each Dt. D-value was estimated using Eq. (12) at the temperature calculated in the node. This temperature was assumed to be
constant during Dt.

Fig. 4. Experimental (exp) and calculated (calc) temperatures in the center of the
cucumber for (a) various temperatures, (b) various cucumber diameters, and (c)
various brine ow rates ( 105 m3/s) while maintaining the other two factors
constant.

2.4. Physical properties and heat transfer coefcients

Physical properties of cucumbers were obtained from Fasina et
al. (2002), who found that variations in physical properties with
temperature were minor. They reported the following values for
cucumbers: an apparent density of 959 kg/m3; a thermal conductivity of 0.61 W/mK; and a specic heat capacity of 4.04 kJ/kg K.
Heat transfer coefcients were estimated using correlations with
the Reynold and Prandt numbers according to Eqs. (10) and (11)
(Perry and Green (1997).
For the cylinder:

293

R.O. Morawicki, M.E. Schmalko / Journal of Food Engineering 107 (2011) 289295

3. Results and discussion

3.1. D-value and z-value of peroxidase

Fig. 5. Estimated average times with 95%-condence limits to achieve a 5-log

reduction of peroxidase for three diameters and three temperatures.

 
kw
hc 0:683
Re0:466 Pr1=3
Dc

10

For the sphere:

 
kw
hs 0:6
Re0:5 Pr1=3
Ds

11

Peroxidase was used as a reference for pasteurization effectiveness because this enzyme is responsible for tissue softening in
cucumbers and other vegetables (Matsui et al., 2008; Polata
et al., 2009; Seyhan et al., 2002; Nicholas and Pug, 1962). The
activity of residual peroxidase for the temperatures studied, which
is expressed as a percentage of the activity at time zero, follows an
approximately linear pattern when plotted against the temperature (with coefcients of determination (R2) between 0.9782 and
0.9984) (Fig. 3). From the regression lines, D-values were calculated for each temperature and presented in Table 1. The z-value,
based on the D-values, was 8.0 0.36 C (mean standard error)
with an MPE of 4.1%. A z-value of 10 C has been reported previously (Nicholas and Pug, 1962). The discrepancy is likely due to
differences in cucumber varieties and methodology. An equation
was also developed to establish the dependence of D-value and
temperature (Eq. (12)), and this equation was used to estimate
the D-value used in Eq. (9).



Tck;t  71:1
D-value a log 3:653 
8

2.5. Statistical analysis

An analysis of variance with StatGraphics (2009) statistical
package was used to determine whether the time to reduce enzyme activity was related to the factors used in this study.
Table 1
D-values at four temperatures for cucumber peroxidase.
T (C)

D-value (s) condence interval

Standard error

71.1
73.9
76.7
79.4

4323 738
2049 200
765 114
390 24

301
82
46
8

12

A similar expression of the relationship of D-value and temperature was used by Matsui et al. (2008) when working with deactivation of peroxidase in coconuts by microwave heating. Other
researchers use an Arrhenius model to describe the relationship
of enzyme activity with temperature, however, the model used
in Eq. (12) has been chosen due to its simplicity.
3.2. Application of the model
There was good agreement between the experimental data and
estimated values in the proposed model with a mean percent error
(MPE) of 4.89% and a root mean square error (RMSE) of 1.29 C (Ta-

Table 2
Experimental design, prediction errors of the model (MPR and RMSE), and treatment time to achieve a 5-logarithmic reduction in the activity of peroxidase (log C0/C = 5) for an
overall treatment of the whole cucumber (Time O) and for treatment at the center of the cucumber (Time C).
Experiment#

Temperature (C)

Flow  105 (m3/s)

Diameter  102 (m)

MPE

RMSE (C)

Time C (s)

Time O (s)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Mean value

76.7
76.7
76.7
76.7
76.7
76.7
76.7
76.7
76.7
79.4
79.4
79.4
79.4
79.4
79.4
79.4
79.4
79.4
82.2
82.2
82.2
82.2
82.2
82.2
82.2
82.2
82.2

6.31
6.31
6.31
12.62
12.62
12.62
18.92
18.92
18.92
6.31
6.31
6.31
12.62
12.62
12.62
18.92
18.92
18.92
6.31
6.31
6.31
12.62
12.62
12.62
18.92
18.92
18.92

2.5
3.3
4.0
2.5
3.3
4.0
2.5
3.3
4.0
2.5
3.3
4.0
2.5
3.3
4.0
2.5
3.3
4.0
2.5
3.3
4.0
2.5
3.3
4.0
2.5
3.3
4.0

6.47
0.01
20.19
2.10
3.12
2.79
9.1
1.39
2.12
0.05
3.39
4.50
4.13
6.12
9.04
2.84
16.22
0.29
1.11
1.11
2.80
2.14
4.11
7.33
11.72
5.49
2.33
4.89

2.03
0.95
1.85
0.99
1.08
0.98
1.91
0.82
0.94
0.39
1.42
1.04
1.41
1.55
1.96
0.84
2.63
1.77
1.29
1.17
1.03
1.11
1.33
1.62
1.73
0.56
0.39
1.29

600
935
1134
632
951
1222
646
926
1170
600
885
1176
593
868
1186
596
860
1140
566
838
1125
564
832
1124
590
827
1120
878

420
570
626
429
588
694
437
568
649
513
538
620
392
626
502
394
497
607
564
362
470
360
465
563
390
461
560
514

R.O. Morawicki, M.E. Schmalko / Journal of Food Engineering 107 (2011) 289295

(a) 90

T media

80

Node 1

70
Temperature (oC)

ble 2). Fig. 4 shows experimental and calculated data for nine
experimental runs during the rst 600 s of heating. Each gure
shows the rise of temperature in the center of the cucumber with
time for one variable factor (brine temperature, ow, or diameter)
while maintaining the other two factors constant. It is important to
mention that initially the cucumbers were at room temperature
which resulted in a variation in the initial cucumber temperatures
of each experimental run. Also, the location of thermocouples was
different for each experiment and cucumber because of the difculty of placing them all at the same distance from the surface.
Fig. 4(a) shows how the temperature was inuenced by three
different brine temperatures (experimental and predicted data)
with a xed brine ow rate (6.31  105 m3/s) and a cucumber
diameter (3.3  102 m). The overlap between the different curves,
mainly in the rst stages of heat treatment, is likely due to the
small difference in the brine temperatures for the different experiments (2.8 C) as well as the different initial cucumber temperatures and the variable position of the thermocouples. Obviously,
higher temperatures would speed up heating, but they also would
increase quality deterioration on the surface. Fig. 4(b) shows the
inuence of diameter on the temperature evolution. As expected,
diameter has an important inuence on the heat penetration
curves. Fig. 4(c) shows the heating curves for three different brine
ows (1, 2, and 3 gal/min). Similarly to Fig. 4(a), no signicant difference in the heating rate was observed over the range of ows
tested.

60

Node 5

50

Node 10

40
30

Node 15

20
10
0

(b) 90

100

200

70

400

500

600

400

500

600

400

500

600

Node 1

60
Node 5

50

Node 10

40
30

Node 15

20

3.3. Inuence of factors

300
Time (s)

T media

80
Temperature (oC)

294

10

3.4. Temperature proles

Fig. 6 shows the calculated temperature proles at nodes 1, 5,
10, and 15 for the three diameters (a) 2.5  102 m, (b)
3.3  102 m, and (c) 4  102 m while ow and brine temperature were parameterized. Data came from experiments 13, 14, and
15 with a brine temperature set at 79.4 C and brine ow at
12.6  105 m3/s.
The large temperature difference observed between nodes 1
and 15 indicates these two nodes are subjected to large differences
in heat treatment as well. When taking into consideration the
cucumber diameters, temperature in the center node was highly

100

200

300
Time (s)

(c) 90

T media

80
Temperature (oC)

The time that it takes to produce a 5-log reduction of the peroxidase activity in the center (node 15) was estimated using the model for the different diameters, ows, and temperatures.
Subsequently, an analysis of variance was conducted, with time
as a dependent variable and the other factors (brine temperature,
diameter, and ow) as independent variables. According to the
analysis, the duration of heating depended on brine temperature
(p < 0.0395) and cucumber diameter (p < 0.0004) but not on the
brine ow (p < 0.9197). Fig. 5 shows the mean times and the 95%
condence intervals for three cucumber diameters and brine temperatures. As expected, the length of pasteurization time increased
with increasing cucumber diameters and decreasing temperatures.
The fact that brine ow had no inuence on heating was due to
a predominant internal heat of the cucumber, which is explained
by the Biot number. The calculated Biot for the three experimental
ows ranged from 5.6 to 11.9, and a Biot above 1 indicates that the
heat ux was dominated by the internal heat resistance (Kreith
and Bohn, 1997). Working with a different system, Fasina and
Fleming (2001) used convective coefcients between 500 and
6000 W/m2K and found that they do not have an inuence on
the internal temperature of the cucumbers. Therefore, the more
modest external heat transfer coefcients in this study, which varied from 212 to 511 W/m2K, are in agreement with the values reported by Fasina and Fleming (2001).

70

Node 1

60
50

Node 5

40

Node 10

30
20

Node 15

10
0
0

100

200

300
Time (s)

Fig. 6. Temperature prole during heating of cucumbers with diameters of

2.5  102 m (a); 3.3  102 m (b); and 4.0  102 m (c), corresponding to experiments 13, 14, and 15, respectively, with the brine temperature of 79.4 C and a ow
of 12.6  105 m3/s.

dependent on the diameter. For instance, to reach 50 C in the center of the cucumber with a 2.5  102 m diameter required 235 s
while the one with a 3.3  102 m diameter required 386 s and
the one with a 4  102 m -diameter required 600 s.
When developing the model, we assumed there was no difference between neighboring nodes of the cylinder and the semispherical ends (lack of a longitudinal temperature prole). To
check this assumption, we compared the temperatures between
the neighboring nodes during heating and found differences greater than 2 C only at the beginning of heating (<120 s). After 120 s,
the differences were less than 1 C, so the assumption was considered acceptable.

R.O. Morawicki, M.E. Schmalko / Journal of Food Engineering 107 (2011) 289295
Table 3
Values of log (C0/C) at selected nodes for an overall 5-logarithmic cycle reduction in
peroxidase activity in the whole cucumber from experiments 12, 13, and 14.
Experiment #

12
13
14

Node
1

10

15 (center)

10.66
12.38
13.41

3.01
2.32
1.72

1.01
0.54
0.28

0.65
0.30
0.14

3.5. Overall treatment vs. center node

The time treatment required for a 5-log reduction in the activity
of peroxidase at the center node (the most unfavorable location)
was compared to the time needed to obtain an overall 5-log reduction in peroxidase activity in the whole pickle (Table 2, Time C
and Time O columns). On average, a 5-log deactivation of peroxidase in the center required a heat treatment that was 71% longer
than the treatment for an overall 5-log reduction activity within
the whole pickle. And when a 5-log treatment was reached in
the center, the overall reduction in peroxidase activity was 18.5logarithmic cycles (log(C0/C) = 18.5).
Either an overall or an in-the-center 5-log reduction in peroxidase activity will causes a gradient of enzyme activity that increases from the periphery to the center, which is caused by the
low thermal conductivity of the cucumber tissue. This is illustrated
in Table 3 with the difference in the concentration of peroxidase
expressed as log(C0/C) for nodes 1, 5, 10, and 15 with the same
overall treatment. For each treatment, there were signicant differences between the surface and the center node with differences up
to two orders of magnitude for the largest diameter.
3.6. Determination of a representative node
Considering that, due to the thermal gradient, each node was
subjected to different heat treatments; we determined the intermediate node between the surface and the center that had the
same heat treatment as the overall treatment. This node was designated a representative node for the overall treatment that separated the nodes with greater enzymatic activity reductions
(towards the surface) from the ones with less enzymatic activity
reductions (towards the center). Again, an overall 5-log reduction
was used and, for 23 experiments, node 3 was considered representative while, for only three experiments, node 2 was the one.
When this was translated into diameter ratios, the representative
distance from the center was located between 0.87 and 0.93 of
the radius.
4. Conclusions
The nite-difference method estimated the internal temperature prole during out-of-container pasteurization of cucumber
fruits with a relatively low error. The time necessary to apply a
determined heat treatment depended on the cucumber diameter
and brine temperature, but not on brine ow. During the heat
treatment, a signicant internal temperature gradient was ob-

295

served, which caused a variable heat treatment in the direction

of the heat ow. For a specic heat treatment, the most unfavorable point, the center, required 71% more time in comparison at
an overall scale. The distance from the center that required the
same heat treatment as the overall treatment was between 0.87
and 0.93 of the diameter.
A 5-logarithmic cycle reduction in peroxidase activity in the
center would subject the whole pickle to an activity reduction of
18.5-logarithmic cycles with quality deterioration of the surface
likely. Therefore, based on the relatively high stock rotation of
pickled products, an overall 5-log reduction of the peroxidase
may prove sufcient. Clearly, a shelf-life study would be needed
to corroborate or disprove this hypothesis.
Acknowledgement
The authors would like to express their appreciation to Dr. Ron
Buescher for his contribution of valuable information used in this
research.
References
Buescher, R.W., McGuire, C., 1986. Peroxidase activities in cucumber pickle
products. Journal of Food Science 51 (4), 10791080.
Breidt, E., Hayes, J.S., Osborne, J.A., McFeeters, R.F., 2005. Determination of 5-Log
pathogen reduction times for heat-processed, acidied vegetable brines. Journal
of Food Protection 68 (2), 305310.
Fasina, O.O., Fleming, H.P., 2001. Heat transfer characteristics of cucumber during
blanching. Journal of Food Engineering 47 (3), 203210.
Fasina, O.O., Fleming, H.P., McFeeters, R.F., 2002. Thermal and and rheological
properties of brine from fermented and sulted-preserved cucumbers. Journal
of Food Process Engineering 25 (4), 307322.
Hoffman, J.D., 1992. Numerical methods for engineering and scientists. McGrawHill, New York, NY (pp. 346352).
Kreith, F., Bohn, M.S., 1997. Principles of Heat Transfer, fth ed. West Publishing
Company, St. Paul, MN (pp. 120).
Lin, Y.L., Li, S.J., Zhu, Y., Bingol, G., Pan, Z., McHugh, T.H., 2009. Heat and mass
transfer modeling of apple slices under simultaneous infrared dry blanching
and dehydration process. Drying Technology 27, 10511059.
Martens, M., Scheerlinck, N., De elie, N., De Baerdemaeker, J., 2001. Numerical model
for the combined simulation of heat and enzyme inactivation kinetics in
cylindrical vegetables. Journal of Food Engineering 47 (3), 185193.
Matsui, K.N., Wilhelms Gut, J.A., Vitoriano de Oliveira, P., Tadini, C.C., 2008.
Inactivation kinetics of polyphenol oxidase and peroxidase in green coconut
water by microwave processing. Journal of Food Engineering 88 (2), 169176.
Mattos, F.R., Fasina, O.O., Reina, L.D., Fleming, H.P., Breidt, F., Damasceno, G.S.,
Passos, E.V., 2005. Heat transfer and microbial kinetics modeling to determine
the location of microorganisms within cucumber fruit. Journal of Food Science
70 (5), E324E330.
Nicholas, R.C., Pug, I.J., 1962. Over- and under pasteurization of fresh cucumber
pickles. Food Technology 16, 104108.
Perry, R.H., Green, D.W., 1997. Perrys Chemical Engineers Handbook, Seventh ed.
McGraw-Hill, New York, NY (pp. 518).
Polata, H., Wilinska, A., Bryjak, J., Polakovic, M., 2009. Thermal inactivation kinetics
of vegetable peroxidases. Journal of Food Engineering 91 (3), 387391.
Sigma, 1995. 3,30 ,5,50 -Tetramethylbenzidine (TMB) liquid substrate system. Sigma
Biosciences, Sigma Chemical Company. St. Louis, MO.
Seyhan, F., Tijskens, L.M., Evranus, O., 2002. Modeling temperature and pH
dependence of lipase and peroxidase activity in Turkish hazelnuts. Journal of
Food Engineering 52 (4), 387395.
StatGraphics, 2009. Centurion XV. StatPoint Technologies Inc., Warrenton, VA.
Wang, Z., Wu, H., Zhao, G., Liao, X., Chen, F., Wu, J., Hu, X., 2007. One-dimensional
nite-difference modeling on temperature history and freezing time of
individual food. Journal of food Engineering 79 (2), 502510.