1642

Journal of Food Protection, Vol. 66, No. 9, 2003, Pages 1642–1649

Copyright q, International Association for Food Protection

Application of Ultrasound-Assisted Thermal Processing for

Preservation and Quality Retention of Liquid Foods
M. ZENKER,* V. HEINZ, AND D. KNORR

Department of Food Biotechnology and Food Process Engineering, Berlin University of Technology, Königin-Luise-Str. 22, D-14195 Berlin, Germany

MS 02-326: Received 19 September 2002/Accepted 15 April 2003

Downloaded from www.jfoodprotection.org by 141.101.201.64 on 10/09/18. For personal use only.

ABSTRACT
A continuously working pilot plant–scale prototype was used to evaluate the effects of continuous-ow ultrasound-
temperature treatment for bacterial decontamination of model suspensions and various liquid food systems such as milk, fruit,
and vegetable juices. Escherichia coli K12 DH 5 a and Lactobacillus acidophilus were used as test microorganisms. In
addition, treated juices were investigated for damage caused by heat or ultrasound-induced degradation of sensory and nutri-
tional properties after treatment and storage. Changes in color and destruction of heat-labile and slightly oxidizable L -ascorbic
acid content were monitored as an index to measure processing effects. Results were assessed with respect to the total energy
requirement and compared with those using a conventional, indirect heating method having similar processing conditions. For
the bacteriological process evaluation, the temperature- and time-dependent process lethality was used as the basis; for the
Journal of Food Protection 2003.66:1642-1649.

quality- and energy-related investigations, the degree of bacterial inactivation was used. At identical degrees of bacterial
inactivation, the ultrasound-assisted thermal treatments required a lower processing temperature than treatment with conven-
tional thermal processing. However, according to energy balances, the total energy consumption was not reduced compared
to conventional heating. Indications for a positive in uence on shelf life, with improvements in surface color stability (lightness)
and L -ascorbic acid retention, were found among quality parameters of treated orange juice.

The concept of ultrasound-assisted thermal processing,
which makes use of the synergy between ultrasound and
heat for bacterial inactivation, has proven to be of potential
interest in food preservation. First introduced by Ordonez
(9), its applicability was predicted for the support of con-
ventional thermal treatments. Suited to pumpable liquids
that can be processed in a continuous- ow arrangement
with clean or aseptic Ž lling (4), the introduction of ultra-
sound was considered to reduce the process temperatures,
the process times, or both of pasteurization or sterilization
processes to achieve the same lethality values as with con-
ventional processes (8). The reduction of the temperature,
the processing time, or both should result in a lower energy
requirement and, above all, in improved food quality.
Several research groups extensively studied the bacte-
ricidal effect of ultrasound, particularly its synergistic po-
tential when applied simultaneously with heat (5, 14). The
in uence of the critical process parameters has been eval-
uated with regard to the treatment efŽ cacy on different
types of bacteria (5, 21). Recent investigations have shown
the in uence of amplitude, external static pressure, and
temperature (10, 12, 13, 14), as well as pH and composition
of treatment medium (11), as the key variables.
However, the majority of studies on bactericidal efŽ -
cacy of ultrasound and ultrasound-assisted thermal process-
ing were investigated in laboratory-scale, batch-type, ultra-
sonic reactors. Detailed experience with continuous- ow
reactors are rare, particularly with reference to energy and
* Author for correspondence. Tel: 149 30 314 71441; Fax: 149 30 832
7663; E-Mail: marco.zenker@tu-berlin.de.
food quality assessment. Recently, some studies have in-
vestigated ultrasound or ultrasound-assisted thermal effects
in continuous  ow (17, 19). The experimental results
achieved in these studies led to further investigations into
the total energy consumption and effect on product quality
of a combination preservation process (18).
To compare the ultrasound-assisted treatment with con-
ventional heating for bacterial inactivation, various nonlin-
ear effects associated with intense sound Ž elds should be
considered, such as ultrasound-generated acoustic stream-
ing or the temperature effect as a result of the speciŽ c ab-
sorption of acoustic energy.
When considering the temperature effect, the temper-
ature-dependent process lethality has to be calculated from
each treatment. A relative pasteurizing capacity can be de-
termined by the conventional F-value approach in accor-
dance with the Ball concept (1, 7). This approach uses the
integrated time-temperature proŽ les of the investigated pro-
cesses, the decimal reduction time at reference temperature
TR , and the z-value. The analysis is based on the assumption
that there is a uniform temperature and velocity distribution
in the radial direction, resulting from a perfectly mixed
product  ow, and that the internal heat generation during
sonication is homogeneous over the radius because of a
volumetric energy dissipation with an exclusively axial gra-
dient.
In the present work, the potential use of the synergistic
effects of ultrasound and temperature in a continuously op-
erating pilot plant unit was studied. An experimental design
was developed to evaluate the effects of ultrasound-assisted
thermal treatment on bacterial inactivation, as well as on
 J. Food Prot., Vol. 66, No. 9
sensory and nutritional properties of liquid foods. The re-
sults were assessed with respect to the energy requirement
and compared with those obtained using a conventional
heating method.
MATERIALS AND METHODS
Bacterial cultures, growth conditions, and assessment of
bacterial viability. Experiments were performed using gram-neg-
ative Escherichia coli K12 DH 5 a (Hygiene Institut, Hamburg,
Germany) and gram-positive Lactobacillus acidophilus ATCC
4356 (DSM, Braunschweig, Germany). Stock cultures were in-
oculated into Standard I nutrient broth (Merck, No. 7882, Darm-
Downloaded from www.jfoodprotection.org by 141.101.201.64 on 10/09/18. For personal use only.
stadt, Germany), in the case of E. coli, or into MRS broth (Difco,
Sparks, Md.), in the case of L. acidophilus, and incubated with
shaking (140 rpm) at 308C for 48 h. After incubation, the cultures
reached the stationary growth phase and attained maximum ther-
motolerance (2). Aliquots from the cell suspensions were sus-
pended in sample medium yielding a Ž nal cell population of 1 3
107 to 1 3 108 CFU per ml.
After each treatment, 0.1 ml of inoculated sample medium
was taken, serially diluted (1:10) with Ringer solution (Merck, no.
15525), and plated by drop plating on ENDO agar (Merck, no.
Journal of Food Protection 2003.66:1642-1649.
4044), in the case of E. coli, or pour plated in Rogosa agar
(Merck, no. 5413), in the case of L. acidophilus (2). The microbial
population of the controls and of the treated suspensions were
determined by counting colony-forming units after incubation at
378C for 48 h.
Laboratory-scale, batch-type treatments. To control the
temperature during conventional and ultrasound-assisted heating,
9 ml of sample medium was placed in a thin glass vessel and
heated by a controlled thermostat bath (Haake, Model DC5, Karls-
ruhe, Germany) until steady state conditions were reached. Inoc-
ulation of 1 ml cell suspension was done when the medium (9 ml
treated volume) had reached the temperature selected for the treat-
ment. In the course of the ultrasound-assisted treatment, the Ž nal
temperature was adjusted during sonication. Ultrasound at 20 kHz
and 50, 110, or 160 mm (equivalent to a power input of 7.2, 15.4,
or 17.6 W) of wave amplitude was applied to the medium using
a Sonopuls HD 2070 homogenizer (Bandelin Electronic and Co.
KG, Berlin, Germany) equipped with an SH 70 G horn and a KE
76 tip. For the accurate speciŽ cation of the sound wave amplitude,
the power control and the sonotrode stroke has been calibrated
with the aid of a laser vibrometer (OFV 3000, Polytec, Berlin).
Pilot-scale treatment in continuous  ow. About 60 liters of
freshly inoculated sample medium (phosphate buffer, ultrahigh-
temperature [UHT] milk, and orange and carrot juice with a Ž nal
cell count of 1 3 106 to 1 3 107) were treated after 1 h adaptation
time at 208C for the test organisms. The experimental set-up em-
ployed has been described previously (21, 22). The sample me-
dium was indirectly heated by a plate heat exchanger and then
pumped through the holding section (diameter, 20 mm; length,
1,500 mm) into the cylindrical sonication cell (inside: diameter,
19.8 mm; height, 380 mm). It was then chilled by countercurrent
cooling after passing through a pressure control valve, which al-
lowed for regulation of the  ow rate and operation in the range
of 0.35 to 0.45 MPa system pressure.
During ultrasound-assisted heating experiments, sonication
was carried out using an industrial ultrasound processor (Model
UIP 1000, Dr. Hielscher GmbH and Co. KG, Teltow, Germany)
attached at the head of the vertically installed sonication cell. The
acoustic source creates a high-frequency oscillation of 19.3 kHz
at amplitudes between 45 and 55 mm within the medium  owing
ULTRASOUND AND HEAT DECONTAMINATION 1643
from below. It was equipped with a specially created conical tip
with an 8.32-cm2 front area, which was calibrated for frequency
resonance between 0.35 and 0.45 MPa to maintain a maximal
power input of approximately 700 to 800 W.
Because of the strong dependence of the maximal sound en-
ergy from the processing temperature and the physical properties
of the sample medium, the input power of the high-frequency
generator was controlled. The data of the supplied power were
simultaneously used to calculate the required electrical energy of
the generator. Moreover, the effective acoustical power was de-
termined after previous measurement of the power loss of the
ultrasound processor. The pilot plant unit was provided with
equipment for permanent monitoring and data acquisition of mass
 ow, processing pressure, and temperature. Among other things,
it allowed the control and determination of mean time-temperature
proŽ les at any operating condition. Inactivation experiments were
performed at volume  ow between 20 and 60 liters/h and highest
output intensity level of the sonicator. During operation, the heat
transmission rate of the heat exchanger was changed in order to
locate the range in which bacterial degradation could be measured.
At regular time intervals after reaching steady state temperature,
1-ml aliquots of sample medium were taken and immediately
cooled in 9 ml of Ringer solution. For comparing the degree of
destruction, D- and z-values were determined, taking into account
the standard deviation and the 95% conŽ dence limit of variations
in slope of Ž tted regression lines.
Process evaluation. The temperature-dependent process le-
thality (PL, s) (pasteurizing unit) was calculated as the cumulative
sum of lethal rate according to equation 1, where T(t) is the av-
erage temperature at a speciŽ c location, TR is the reference tem-
perature, t is time, and z is the temperature dependency of the
kinetic reaction rate of the test organisms used.
E
t1
T(t) 2 T R
PL 5 10 dt (1)
t0 z
PL from the treatments evaluated is the time required to get the
same lethal effect as obtained at reference temperature. The kinetic
parameter z was derived from the batch-type experiments using
the thermal death time method from the semilogarithmic plot of
D-value versus time curves in the same manner as for the previous
determination, the decimal reaction time D from the semilogarith-
mic plot of survivor ratio versus time curves. Over the investi-
gated temperature range (E. coli: temperature by conventional
heating [T] 5 56 to 68, ultrasound-assisted temperature [TS] 5
48 to 60; L. acidophilus: T and TS 5 52 to 60), z was constant.
Energy efŽ ciency analysis. The total energy consumption
was calculated for the ideal loss-free process and under the spe-
ciŽ c operating conditions of the pilot plant unit. The calculation
was made for a temperature increment of DT with the equation
for heating power, the enthalpy balance of the heat transfer unity,
and the measured effective acoustic power output of the ultrasonic
source (equation 2). The results were determined as speciŽ c values
with equal degrees of bacterial inactivation as the basis.
E S 5 (PowI 2 Pow L )t S (2)
ES (kJ/kg) is the energy input per unit mass, tS (s/kg) is time per
mass, PowI (kW) is the input or supplied power, and PowL (kW)
is the power loss measured as power input during operation of the
ultrasonic source in air.
Quality assessment. For posttreatment investigations, pas-
teurization was carried out in laboratory-scale batch equipment at
758C and different treatment times (1 to 30 s) or speciŽ c sound
 1644 ZENKER ET AL.
Downloaded from www.jfoodprotection.org by 141.101.201.64 on 10/09/18. For personal use only.
FIGURE 1. Survival (A) and thermal dead time curves of E. coli
K12 DH 5 a and thermal dead time curves of L. acidophilus after
Journal of Food Protection 2003.66:1642-1649.
ultrasound (V), thermal (h), and ultrasound-assisted thermal
treatments (v, sound wave amplitude 5 110 mm) in phosphate
buffer (pH 7).
energy inputs (43.2 and 1,296 kJ/kg). To study storage behavior,
juices were treated continuously at the pilot plant scale at PL60 5
360 (thermal) and 220 s (ultrasound-assisted thermal treatment)
and at a speciŽ c sound energy input of 96 to 110 KJ/kg during
sonication.
L -ascorbic acid levels were determined by the colorimetric
assay from Boehringer (No. 409 677; Mannheim, Germany) and
carried out by adding 0.1 ml of the Ž ltered sample into cuvettes
at room temperature, 1 ml of 0.75 mM 3-(4,5-dimethylthiaolyl-
2)-2,5-diphenyltetrazolium bromide solution (MTT), and 1.5 ml
distilled water. In the presence of 0.1 ml of 15 mM 5-methyl-
phenazine methosulfate at pH 3.5, the MTT was reduced by the
ascorbic acid. The resulting formazan was photometric and was
detected by absorbance at 578 nm after 15 min incubation at 378C.
Color after treatment or during the storage period was in-
strumentally determined using the CR 200 model Minolta Chro-
mameter (Minolta Camera Co., Osaka, Japan) with triplicate mea-
surements for each determination. The CIE L*a*b* color notation
system was applied to measure the parameters L* (lightness), a*
(color axis from red to green), and b* (yellow-blue axis). The net
color difference (DE) was calculated from the color parameters as
follows, where the subscript 0 indicates the initial color.
DE 5 Ï (L 2 L0)2 1 (a 2 a0)2 1 (b 2 b0)2 (3)
TABLE 1. Kinetic factors D60 and z48–68 for E. coli DH 5 a and L. acidophilus inactivation by thermal and ultrasound-assisted thermal
treatments (sound wave amplitude 5 110 mm) in the batch-type reactor
E. coli K 12 DH 5 a
Phosphate buffer Carrot juice
(pH 7.0) (pH 5.9)
Ta TS T TS
D60-value (s) 84.6 23.1 84.3 23.9
z48–68-value (8C) 6.9 16.3 8.7 15.3
a T, conventional heating; TS, ultrasound-assisted treatment.
J. Food Prot., Vol. 66, No. 9
RESULTS AND DISCUSSION
Laboratory-scale, batch-type experiments. The de-
struction kinetics of the test organisms for both conven-
tional and ultrasound-assisted heating are best described by
a Ž rst-order reaction, which enabled the determination of
D- and z-values by linear regression, as well as the evalu-
ation of the inactivation results by using a simple time law
(Fig. 1A). The kinetics for sonication at ambient tempera-
ture showed the same characteristics.
On the basis of batch-mode kinetics data (Table 1), a
signiŽ cant enhanced effect on E. coli and L. acidophilus
destruction as a result of the ultrasound-assisted thermal
processing was detected. In buffer media at 608C, the com-
bination process reduced the D-value of E. coli by 72.7%
and of L. acidophilus by 38.6%, compared to the D-value
of the heat treatment alone, and by 71.6 or 70.1% (E. coli)
and 34.2% (L. acidophilus) when tested in carrot juice,
milk, or orange juice. Accordingly, a 3.5-fold (E. coli) and
a 1.5-fold (L. acidophilus) reduction in process time or, at
the same D-value, a 1.13-fold (E. coli) or 1.04-fold (L. ac-
idophilus) reduction in temperature could be carried out to
inactivate identical microbial loads.
Moreover, the inactivation rate of the ultrasound-as-
sisted thermal treatment was greater than that of heat added
to the inactivation rate of ultrasonication when applied sep-
arately. This synergistic effect was detected in the temper-
ature range of 48 to 678C for E. coli and 52 to 608C for L.
acidophilus. For ultrasound treatment at ambient tempera-
ture, D28-values of 170.2 s for E. coli and 1,109 s for L.
acidophilus was detected.
From thermal death time curves (Fig. 1), it can be seen
that the cell sensitivity of both E. coli and L. acidophilus
to the ultrasound-assisted heat treatment differed widely in
spite of an almost identical heat resistance. This might re-
sult from the speciŽ c effect of ultrasound on the cell wall
and the differences in the cell wall structures between the
species investigated. According to the literature, differences
in cell sensitivity could be the result of the more tightly
adherent layer of peptidoglycans in gram-positive cells and
the cell shape—in particular, the relationship of cell surface
and volume (4, 15). Because of the limiting effect of the
vapor pressure and the reduction in  uid tensile strength,
which can cushion the collapse of bubbles and strength of
cavitation (4), the synergism of ultrasound-assisted thermal
treatment signiŽ cantly diminished as the temperature of the
treatment was increased. When approaching the range of
L. acidophilus
UHT milk Phosphate buffer Orange juice
(pH 6.7) (pH 7.0) (pH 3.7)
T TS T TS T TS
77.0 23.0 70.5 43.3 47.3 31.1
7.1 13.6 6.1 7.5 5.5 7.0
 J. Food Prot., Vol. 66, No. 9 ULTRASOUND AND HEAT DECONTAMINATION 1645
Downloaded from www.jfoodprotection.org by 141.101.201.64 on 10/09/18. For personal use only.

FIGURE 2. Effect of ultrasound wave amplitude (0, 50, 110, 160 FIGURE 3. In uence of the treatment media (milk: pH 6.7, fat
mm) on the relationship between D-value and temperature used 3.5%; carrot juice: pH 5.9, 5.3 8Brix; orange juice: pH 3.7, 11.2
Journal of Food Protection 2003.66:1642-1649.

in the ultrasound-assisted thermal treatment of E. coli K12 DH 5 8Brix) on the resistance of E. coli DH 5 a (A) and L. acidophilus
a in phosphate buffer (pH 7). (B) to thermal and ultrasound-assisted thermal treatments (sound
wave amplitude 5 110 mm) in the batch-mode system.

65 to 678C, no signiŽ cant difference was found in lethality

between the thermal and the ultrasound-assisted treatment. formed at volume  ows higher than 28 liters/h (data not
The lethality of the ultrasound-assisted thermal treat- shown). But in the  ow range between 20 and 26 liters/h,
ment was highly in uenced by the amplitude of the sound a consistently higher bacterial inactivation could be detect-
wave (Fig. 2). At a constant temperature of 608C, the D- ed in the ultrasound-assisted thermal treatment (Fig. 4). The
value decreased exponentially from 39.4 to 19.4 s when the observed increased lethality in the combination treatment
sound wave amplitude was increased from 50 to 160 mm.
As suggested by Berliner (3) and Suslick (16), an increase
in amplitude produces an increase in the effective zone of
the liquid and also the range of bubble size undergoing
cavitation. The increase of bacterial lethality is then de-
scribed as the immediate consequence of the higher inten-
sity of a single bubble implosion and the higher density of
cavitation events.
Results on the in uence of treatment media on the le-
thality of thermal and ultrasound-assisted thermal treat-
ments are shown in Figure 3. The effects of the medium
on the inactivation of E. coli were negligible, either by heat
or ultrasound-assisted thermal treatment. However, in the
case of ultrasound-assisted thermal treatment, differences
occurred at lower temperatures, with some higher D-values
in milk and carrot juice compared to the model media (Figs.
1 and 3). With L. acidophilus inactivation in orange juice,
a signiŽ cant decline in resistance to thermal processing, as
FIGURE 4. Inactivation of E. coli DH 5 a (A) and L. aci-
well as to ultrasound-assisted thermal processing, in com- dophilus (B) after conventional (V) and ultrasound-assisted
parison to the inactivation in buffer media was established heating (v) within the pilot plant unit. The x-axis value cor-
(Figs. 1 and 3). The highest D-value reduction occurred at responds to the treatment time required for pasteurization at
the maximum temperature of the investigated temperature the reference temperature of 608C in order to achieve the effect
range and amounted to 32.9% for the thermal and 28.2% shown. Each open and a closed circle represents a treatment
for the combined treatment. with another thermal energy input. Mean residence time within
the ultrasound reaction chamber, ;60 s; speciŽ c sound energy
Pilot-scale experiments in continuous  ow. In a com- input, 96 to 110 kJ/kg; volume  ow, 26 liters/h; phosphate buff-
parison of thermal and ultrasound-assisted treatment, hardly er treatment medium, pH 7; standard deviation of a single ran-
any difference in the inactivation rate of either E. coli or dom sample (six repetitions, log N/No), 0.59 (E. coli) and 0.55
L. acidophilus was observed when treatments were per- (L. acidophilus).
 1646 ZENKER ET AL.
Downloaded from www.jfoodprotection.org by 141.101.201.64 on 10/09/18. For personal use only.
FIGURE 5. Comparison of total energy requirement for E. coli
(A) and L. acidophilus (B) inactivation during thermal and ul-
trasound-assisted thermal treatments at different process inten-
sities. Phosphate buffer treatment media, pH 7; volume  ow, 26
liters/h.
Journal of Food Protection 2003.66:1642-1649.
was independent of the treatment media in which the test
organisms were suspended.
A linear relationship also was detected for the destruc-
tion kinetics (Fig. 4). Moreover, the prediction of bacterial
destruction by thermal treatment (results of laboratory-
scale, batch-type experiments) showed no clear agreement
with the results obtained in continuous  ow. Because of
residence time variation, lower inactivation for the experi-
mental data in the present example (20% for E. coli and
7% for L. acidophilus) occurred simultaneously with the
highest deviation in the slope or the D- or z-value. The
mean regression line was obtained from three consecutive
experiments in the  ow range between 20 and 26 liters/h
(r0 5 0.96 [E. coli], or 0.97 [L. acidophilus]; 95% conŽ -
dence limit of the slope, 7.73 to 9.83 3 102 3 [E. coli] or
1.46 to 1.22 3 102 2 [L. acidophilus]).
SigniŽ cant differences in the lethality of the thermal
and ultrasound-assisted thermal treatments were detectable
in spite of the great deviation for E. coli. During continu-
ous- ow operation at 26 liters/h, the isothermal process at
608C had a 64.0 6 21.8% (E. coli) or a 41.6 6 12.8% (L.
acidophilus) reduction in the D- or P-value. This would
allow at least a 1.7-fold (4.4 6 2.7, E. coli) to 1.4-fold (1.8
6 0.4, L. acidophilus) reduction of the process time or a
1.07-fold (1.09 6 0.02, E. coli) to 1.03-fold (1.04 6 0.01,
L. acidophilus) reduction in the process temperature.
Figure 5 shows the total energy requirements of ther-
mal and ultrasound-assisted thermal treatments for the same
extent of microbial inactivation. Generally, the ultrasound-
assisted process required a higher total energy input in spite
of the strong synergistic effect. The approximately 32 to
36% (E. coli) or 42 to 50% (L. acidophilus) higher total
energy requirement and, simultaneously, a 28 to 31% (E.
coli) or 15 to 24% (L. acidophilus) average lowering of
heat energy consumption—depending upon the process
mode, intensity, and test microorganism (Fig. 5)—are at-
J. Food Prot., Vol. 66, No. 9
FIGURE 6. Spatially presented changes in CIE a*, b*, and L*
values in thermal (A) and ultrasound-assisted thermal treat-
ments of orange juice (B) at 758C and at different process times
(1 to 30 s). SpeciŽ c sound energy input during sonication varied
between 43.2 and 1,296 kJ/kg.
tributed to the substantial electric energy consumption used
for the ultrasound generation necessary to achieve a com-
bined bacterial inactivation of between 1 and 4 log cycles.
For the ideal loss-free process, the total speciŽ c net energy
input was quantiŽ ed as 167.0 to 188.4 kJ/kg (E. coli) or
159.0 to 175.8 kJ/kg (L. acidophilus) for the conventional
thermal process and 227.0 to 247.8 kJ/kg (E. coli) or 239.0
to 256.0 kJ/kg (L. acidophilus) for the ultrasound-assisted
thermal treatment.
When comparing the ideal loss-free process with the
process under operating conditions in the pilot plant, an
approximately 15% higher energy requirement was mea-
sured. The possibility of heat recovery for both processes
was not considered. Even though most of the applied elec-
trical energy was transformed into sound energy (efŽ cacy
.95%) and, afterwards, into heat energy during the ultra-
sound-assisted treatment, no substantial change in the re-
lationships is expected.
Directly after processing in both thermal and combined
treatment of juice samples, visual and quantitatively de-
tectable differences in color could be established (Fig. 6A
and 6B). With increasing process intensity, all three color
parameters (CIE L*, a*, b*) changed, causing a color shift
 J. Food Prot., Vol. 66, No. 9
Downloaded from www.jfoodprotection.org by 141.101.201.64 on 10/09/18. For personal use only.
FIGURE 7. Color quality change of orange juice after thermal (V) and ultrasound-assisted thermal treatments (v) in the pilot plant
and subsequent 49-d storage at 208C.
Journal of Food Protection 2003.66:1642-1649.
toward lighter and brighter. In conventionally treated sam-
ples, the change of color values proceeds from a* 5 25.26,
b* 5 25.75, and L* 5 47.22 to a* 5 26.52, b* 5 27.56,
and L* 5 48.47; in samples from the combined treatments,
the color values progress from a* 5 25.26, b* 5 25.75,
and L* 5 47.22 to a* 5 27.44, b* 5 25.20, and L* 5
48.49. The total color difference, DE, was calculated with
2.53 for the thermal and 2.56 for the ultrasound combined
treatment. Although changes in color values a* and b* can-
not be explained, the increase in lightness (L*) can probably
FIGURE 8. Degradation of native ascorbic acid in orange juice
after thermal (V) and ultrasound-assisted thermal treatments
(v) and storage in bottles at 208C in darkness. In the study, the
rehydrated juice (11.2 8Bx, pH 3.7) was pasteurized with regard
to the L. acidophilus inactivation at PL 60 5 360 s (thermal treat-
ment) and 220 s (ultrasound-assisted thermal treatment) and at
a speciŽ c sound energy input of 96 to 110 kJ/kg during soni-
cation. Means and standard deviations were established from
three replications.
ULTRASOUND AND HEAT DECONTAMINATION 1647
be attributed to partial precipitation of suspended, insoluble
particles in the juice (6).
When comparing thermal and ultrasound-assisted,
heat-treated juice samples, though, changes in the color val-
ues occurred at increasing process time in the same manner,
and an ultrasound in uence was not found. SigniŽ cant dif-
ferences (P , 0.05), except for b* values, between samples
treated at the lowest and highest intensity levels could be
detected.
In the posttreatment samples, no differences in the con-
tent of L-ascorbic acid were apparent. The initial content,
which ranged from 255 to 265 mg/liter, was comparable to
reported data, and the high thermal stability was expected
because of the low pH level of 3.7.
During a 49-d storage period of continuously treated
juice samples, changes in all three color parameters—L*,
a*, and b*—were measured (Fig. 7). Here, the change took
place toward lower values, decreasing at least from 1.0 to
1.5, without exception.
Comparing thermal and ultrasound-assisted treatments
of juice samples, only slight differences for the color values
a* and b* were found. It is interesting to note that lightness
(L*) did not changed at the same rate, and values were
signiŽ cantly (P , 0.05) diminished already after 35 d of
storage. Corresponding to these results, a distinct effect on
the retention of L-ascorbic acid in equivalently treated sam-
ples was detected.
During the Ž rst days after bottling the juice samples,
no signiŽ cant change in L-ascorbic acid content occurred
(Fig. 8). However, at ambient temperatures (208C) and stor-
age in glass bottles, a relatively high monthly loss was
found. As can be seen in Figure 8, the quantity varies be-
cause of the applied treatment. In conventionally heated
juice after 35 d, a signiŽ cant (P , 0.05) reduction of ap-
proximately 6.6% was detected.
The differences in degradation between the treatments
cannot be related to the in uence of pH or light or the
occurrence of metal ions or protection substances, but the
 1648 ZENKER ET AL.
degradation seems restricted to enzymatic or chemical re-
actions. Furthermore, it seems reasonable to assume that
such behavior cannot be attributed to the lower thermal
processing intensity of the combined treatment. It is more
probable that this quality retention was the result of the
extremely effective degassing of the juice by ultrasound.
Any dissolved gases in the liquid or medium act as nuclei
for the formation of cavitation bubbles, which will, because
of the gas content, not easily collapse in the compression
cycle but continue to grow on further rarefaction cycles and
 oat to the surface. As a result, the stability of L-ascorbic
acid could be affected by lowering the oxidative decay re-
Downloaded from www.jfoodprotection.org by 141.101.201.64 on 10/09/18. For personal use only.
taining its native conŽ guration.
CONCLUSIONS
In process evaluation for bacterial inactivation, the cri-
terion of thermal equality was the basis for the accurate
comparison of conventional and ultrasound-assisted treat-
ments. This involved the integration of the experimentally
measured time-temperature proŽ les of the pilot plant unit,
which was performed on the assumption of a uniform radial
Journal of Food Protection 2003.66:1642-1649.
temperature and velocity distribution. As a Ž rst step of pro-
cess evaluation, the temperature- and time-dependent pro-
cess lethality was calculated without consideration of the
spatial temperature and residence time distribution. Rather,
in accordance with the low measured radial temperature
gradients in the holding and sonication sections, as well as
the comparable low spread of the corresponding residence
time proŽ les, lethality was calculated from a point location
temperature, and the mean residence time was deŽ ned as
the ratio between the volume of the corresponding test sec-
tion and the  ow rate.
Applying the F-value approach and using the experi-
mental conditions, the ultrasound-assisted thermal treatment
has proven to be suited for achieving a more efŽ cient bac-
terial decontamination of the model suspension, as well as
of the liquid food systems compared to the conventional
indirect heating method. During continuous- ow operations
at 26 liters/h with an isothermal process at 608C, a 64.0 6
21.8% (E. coli) or a 41.6 6 12.8% (L. acidophilus) reduc-
tion in D- or P-value was obtained, which relates to a re-
duced process time of at least 1.7-fold (E. coli) to 1.4-fold
(L. acidophilus) or a process temperature of 1.07-fold (E.
coli) to 1.03-fold (L. acidophilus). However, it could be not
conŽ rmed that the combined process leads to a lowering of
the energy requirement. Because of the high amount of
electrical energy required for ultrasound generation, a high-
er total energy consumption resulted.
Encouraging results were obtained from the quality as-
sessment on ultrasound-assisted, treated orange juice be-
cause no ultrasound-related detrimental effects on the qual-
ity attributes that were investigated were found. After 35 d
of storage, a signiŽ cant (P , 0.05) enhanced stability of
surface color (lightness) as well as a signiŽ cant (P , 0.05)
lower decay of L-ascorbic acid concentration could be es-
tablished. An ultrasound-induced agglomeration and a de-
gassing effect were proposed as the reason for this in-
creased stability.
From these Ž ndings, it is suggested that, for some
J. Food Prot., Vol. 66, No. 9
food products, ultrasound-supported thermal pasteuriza-
tion might offer a way to extend microbial, sensory, and
nutritional shelf life. With this simultaneous application,
no direct energy saving, but a shift to a more comfortable
and readily assessable energy form, can be achieved.
ACKNOWLEDGMENTS
This research project was supported by the FEI (Forschungskreis der
Ernährungsindustrie e.V., Bonn), the AiF, and the Ministry of Economics
and Technology (project no. 11 395 N).
REFERENCES
1. Ball, C. O. and F. C. W. Olson. 1957. Sterilization in food technol-
ogy—theory, practice and calculations. McGraw-Hill, New York.
2. Baumgart, J. 1996. Mikrobiologische Untersuchung von Lebensmit-
teln. Behr’s Verlag, Hamburg.
3. Berliner, S. 1984. Application of ultrasonic processors. Int. Biotech-
nol. Lab. 2:42–49.
4. Earnshaw, R. G. 1998. Ultrasound: a new opportunity for food pres-
ervation, p. 183–192. In M. J. W. Povey and T. J. Mason (ed.),
Ultrasound in food processing. Blackie Academic & Professional,
London.
5. Garcia, M. L., J. Burgos, B. Sanz, and J. A. Ordonez. 1989. Effect
of heat and ultrasonic waves on the survival of two strains of Ba-
cillus subtilis. J. Appl. Bacteriol. 67:619–628.
6. Genovese, D. B., M. P. Elustundo, and J. E. Lozano. 1997. Colour
and cloud stabilization in cloudy apple juice by steam heating during
crushing. J. Food Sci. 62:1171–1175.
7. Holdsworth, S. D. 1992. Aseptic processing and packaging of
food products. Elsevier Applied Food Science, London and New
York.
8. Mason, T. J., L. Paniwnyk, and J. P. Lorimer. 1996. The uses of
ultrasound in food technology. Ultrason. Sonochem. 3:253–260.
9. Ordonez, J. A., M. A. Aguilera, M. L. Garcia, and B. Sanz. 1987.
Effect of combined ultrasonic and heat treatment (thermoultrasoni-
cation) on the survival of a strain of Staphylococcus aureus. J. Dairy
Res. 54:61–67.
10. Pagán, R., P. Mañas, I. Alvarez, and S. Condón. 1999. Resistance of
Listeria monocytogenes to ultrasonic waves under pressure at sub-
lethal (manosonication) and lethal (manothermosonication) temper-
atures. J. Food Microbiol. 16:139–148.
11. Pagán, R., P. Mañas, J. Raso, and S. Condón. 1999. Bacterial resis-
tance to ultrasonic waves under pressure at nonlethal (manosonica-
tion) and lethal (manothermosonication) temperatures. Appl. Envi-
ron. Microbiol. 65:297–300.
12. Raso, J., R. Pagán, S. Condon, and F. J. Sala. 1998. In uence of
temperature and pressure on the lethality of ultrasound. Appl. Envi-
ron. Microbiol. 64:465–471.
13. Raso, J., A. Palop, R. Pagán, and S. Condón. 1998. Inactivation of
Bacillus subtilis spores by combining ultrasonic waves under pres-
sure and mild heat treatment. J. Appl. Microbiol. 85:849–854.
14. Sala, F. J., J. Burgos, S. Condón, P. López, and J. Raso. 1995. Effect
of heat and ultrasound on microorganisms and enzymes, p. 176–204.
In G. W. Gould (ed.), New methods of food preservation. Blackie
Academic & Professional, London.
15. Scherba, G., R. M. Weigel, and J. R. O’Brien. 1991. Quantitive as-
sessment of the germicidal efŽ cacy of ultrasonic energy. Appl. En-
viron. Microbiol. 57:2079–2084.
16. Suslick, K. S. 1988. Homogeneous sonochemistry. In K. S. Suslick
(ed.), Ultrasound. It’s chemical, physical and biological effects.
VCH, New York.
17. Villamiel, M., and P. de Jong. 2000. Inactivation of Pseudomonas
 uorescens and Streptococcus thermophilus in trypticase soy broth
and total bacteria in milk by continuous- ow ultrasonic treatment
and conventional heating. J. Food Eng. 45:171–179.
18. Villamiel, M., E. H. van Hamersveld, and P. de Jong. 1999. Review:
effect of ultrasound processing on the quality of dairy products.
Milchwissenschafts. 54:69–73.
19. Williams, A., C. Leadley, E. Lloyd, G. Betts, R. Oakley, and M.
 J. Food Prot., Vol. 66, No. 9
Gonzales. 1999. Ultrasonically enhanced heat transfer and microbi-
ological inactivation. Poster presentation. European Conference on
Emerging Food Science and Technology, Tampere, Finland.
20. Wrigley, D. M., and N. G. Llorca. 1992. Decrease of Salmonella
typhimurium in skim milk and egg by heat and ultrasonic wave treat-
ment. J. Food Prot. 55:678–680.
21. Zenker, M., V. Heinz, and D. Knorr. 1999. Einsatz akustisch erzeug-
Downloaded from www.jfoodprotection.org by 141.101.201.64 on 10/09/18. For personal use only.
Journal of Food Protection 2003.66:1642-1649.
ULTRASOUND AND HEAT DECONTAMINATION 1649
ter Wechseldrücke bei der direkten HTST- und UHT-Erhitzung  üs-
siger Lebensmittel. Lebensm-verfahr.-Verpackungstech. 44(4):189–
192.
22. Zenker, M., V. Heinz, and D. Knorr. 2000. Ultraschall-Temperatur-
Druck-Kurzzeitbehandlung unter Einbeziehung der DirektdampŽ n-
jektion zur Pasteurisierung/Sterilisierung von Lebensmitteln. Final
Report AiF-FV 11395 N.